RU2818968C2 - Method and apparatus for encoding/decoding image and recording medium on which bit stream is stored - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: invention relates to a method and an apparatus for encoding/decoding an image and a recording medium storing a bit stream based on a candidate list and an intra-prediction mode list. Disclosed is a method of encoding/decoding an image, which comprises deriving a first candidate list of a current block using motion information of a neighboring block of the current block to derive a first candidate list for the current block, deriving a second candidate list for the current block using previously recovered motion information, deriving a third candidate list using the first candidate list and a second candidate list and generating a prediction block for the current block using the third candidate list.

EFFECT: higher efficiency of compression.

14 cl, 73 dwg, 5 tbl

Description

Technical field

[0001] The present invention relates to a method and apparatus for encoding/decoding an image and a recording medium storing a bit stream. More specifically, the present invention relates to a method and apparatus for encoding/decoding an image and a recording medium storing a bit stream based on a candidate list and an intra-prediction mode list.

Prior Art

[0002] Recently, demands for high resolution and high quality images, such as high definition (HD) images and ultra high definition (UHD) images, have increased in various application fields. However, higher resolution and quality image data have increased data volumes compared to regular image data. Therefore, when transmitting image data using a medium such as traditional wired and wireless broadband networks, or storing image data using traditional storage media, transmission and storage costs increase. To overcome these problems that arise with increasing resolution and quality of image data, highly efficient image encoding/decoding techniques are required for higher resolution and higher quality images.

[0003] Image compression technology includes various methods, including: an inter-prediction method by predicting a pixel value included in the current picture from a previous or subsequent picture relative to the current picture; an intra-prediction method by predicting a pixel value included in the current picture using pixel information in the current picture; conversion and quantization method for compressing the energy of the residual signal; entropy encoding method by assigning a short code to a value with a high frequency of occurrence and assigning a long code to a value with a low frequency of occurrence, etc. The image data can be efficiently compressed using such image compression technology and can be transmitted or stored.

Disclosure

Technical problem

[0004] An object of the present invention is to provide an image encoding/decoding method and apparatus having improved compression efficiency, and a recording medium storing a bit stream generated by the image encoding/decoding method and apparatus.

[0005] It is further an object of the present invention to provide a method and apparatus for encoding/decoding an image and a recording medium storing a bitstream that has improved compression efficiency using a candidate list.

[0006] Another object of the present invention is to provide a method and apparatus for encoding/decoding an image and a recording medium storing a bit stream that has improved compression efficiency using a list of intra-prediction modes.

Technical solution

[0007] According to the present invention, an image decoding method comprises deriving a first candidate list of a current block using motion information of an adjacent block of the current block to output a first candidate list for the current block, deriving a second candidate list for the current block using previously recovered motion information , deriving a third candidate list using the first candidate list and a second candidate list, and generating a prediction block for the current block using the third candidate list.

[0008] In a video decoding method, a neighboring block includes at least one of a spatially neighboring block and a temporally neighboring block.

[0009] In a video decoding method, deriving a second candidate list includes adding previously recovered motion information to the second candidate list, wherein the previously recovered motion information is motion information of an encoding block decoded immediately before decoding the current block.

[0010] The video decoding method further comprises initializing a second candidate list, wherein the second candidate list is initialized based on a line boundary of a coding tree unit (CTU).

[0011] The video decoding method further comprises, when the number of pieces of motion information included in the second candidate list is a predetermined value, removing the motion information included in the second candidate list of the first among the pieces of motion information included in the second candidate list.

[0012] In the video decoding method, the previously recovered motion information is added in the following order of the motion information included in the second candidate list last.

[0013] The video decoding method further comprises, when the same motion information as the previously recovered motion information is already included in the second candidate list, removing the same motion information from the second candidate list.

[0014] In the video decoding method, the previously recovered motion information is added in the following order of the motion information included in the second candidate list last.

[0015] In the video decoding method, the maximum number of pieces of motion information that can be included in the second candidate list is predetermined.

[0016] The video decoding method further comprises deriving average motion information by calculating an average value of a plurality of pieces of motion information included in the third candidate list, and adding the average motion information to the third candidate list.

[0017] The video decoding method further comprises dividing a current block into a first region and a second region, wherein generating a prediction block for the current block includes generating prediction blocks for the first region and the second region using the third candidate list.

[0018] In the video decoding method, the third candidate list is used to generate an intra-block copy (IBC) prediction block for the current block.

[0019] According to the present invention, an image decoding method comprises determining motion information of a current block, deriving a first candidate list for the current block using motion information of an adjacent block of the current block, deriving a second candidate list for the current block using previously encoded motion information, and deriving a third list of candidates using the first list of candidates and the second list of candidates.

[0020] In a video encoding method, deriving a second candidate list includes adding previously encoded motion information to the second candidate list, wherein the previously encoded motion information is motion information of an encoding block encoded immediately before encoding the current block.

[0021] The video encoding method further comprises initializing a second candidate list, wherein the second candidate list is initialized based on a line boundary of a coding tree unit (CTU).

[0022] The video decoding method further comprises, when the number of motion information pieces included in the second candidate list is a predetermined value, removing the motion information included in the second candidate list first among the motion information pieces included in the second candidate list.

[0023] The video decoding method further comprises, when the same motion information as the previously encoded motion information is already included in the second candidate list, removing the same motion information from the second candidate list.

[0024] In the video encoding method, the number of pieces of motion information that can be included in the second candidate list is predetermined.

[0025] According to the present invention, a computer-readable recording medium stores a bit stream that is received by an image decoding apparatus and used to reconstruct a current block included in the current picture, wherein the bit stream includes prediction information of the current block, the prediction information is used to output a first candidate list for the current block using the motion information of a neighboring block of the current block, the prediction information is used to output a second candidate list for the current block using the previously recovered motion information, and the first candidate list and the second candidate list are used to output a third candidate list used to generate a prediction block for the current block.

BENEFITIVE EFFECTS

[0026] According to the present invention, it is possible to provide an image encoding/decoding method and apparatus having improved compression efficiency, and a recording medium storing a bit stream generated by the image encoding/decoding method and apparatus.

[0027] In addition, according to the present invention, it is possible to provide a method and apparatus for encoding/decoding an image and a recording medium storing a bit stream that has improved compression efficiency using a candidate list.

[0028] In addition, according to the present invention, it is possible to provide a method and apparatus for encoding/decoding an image and a recording medium storing a bitstream that has improved compression efficiency using a list of intra-prediction modes.

DESCRIPTION OF DRAWINGS

[0029] FIG. 1 is a block diagram showing the configuration of an encoding apparatus to which the present invention is applicable.

[0030] FIG. 2 is a block diagram showing the configuration of a decoding apparatus to which the present invention is applicable.

[0031] FIG. 3 is a view schematically showing a partitioning structure in image encoding and decoding.

[0032] FIG. 4 is a view showing an example of intra-prediction.

[0033] FIG. 5 is a view showing an example of inter-prediction.

[0034] FIG. 6 is a view showing an example of transformation and quantization.

[0035] FIG. 7 is a view showing reference samples that are useful for intra-prediction.

[0036] FIG. 8 is a diagram illustrating an upper prediction unit (UPU) used in some embodiments of the present invention.

[0037] FIG. 9 is a diagram illustrating the structure of an encoder according to an embodiment of the present invention.

[0038] FIG. 10 is a diagram illustrating the structure of a decoder according to an embodiment of the present invention.

[0039] FIG. 11 is a flowchart illustrating an encoding/decoding method in accordance with an embodiment of the present invention.

[0040] FIG. 12 is a diagram illustrating the structure of an encoder and decoder in accordance with another embodiment of the present invention.

[0041] FIG. 13 is a diagram illustrating a list of candidates in accordance with an embodiment of the present invention.

[0042] FIG. 14 is a diagram illustrating a method for adding motion information to a candidate list in accordance with one embodiment of the present invention.

[0043] FIG. 15 is another diagram illustrating a method for adding motion information to a candidate list in accordance with one embodiment of the present invention.

[0044] FIG. 16 is a diagram illustrating a method for adding motion information to a candidate list in accordance with one embodiment of the present invention.

[0045] FIG. 17-20 are diagrams illustrating syntax structure in accordance with some embodiments of the present invention.

[0046] FIG. 21 and 22 are diagrams illustrating flowcharts of an encoder and decoder for approximating motion information in accordance with some embodiments of the present invention.

[0047] FIG. 23 and 24 are diagrams illustrating the structure of an encoder and decoder for approximating motion information, in accordance with some embodiments of the present invention.

[0048] FIG. 25 and 26 are diagrams illustrating a method for calculating candidate list potential in accordance with some embodiments of the present invention.

[0049] FIG. 27 and 28 are diagrams illustrating the structure of an encoder and decoder for computing candidate list potential in accordance with some embodiments of the present invention.

[0050] FIG. 29 is a diagram illustrating the structure of an encoder and decoder for calculating candidate list prediction potential in accordance with some embodiments of the present invention.

[0051] FIG. 30 is a diagram illustrating generation of a mixed candidate list in accordance with an embodiment of the present invention.

[0052] FIG. 31 and 32 are diagrams illustrating a method for initializing a candidate list in accordance with some embodiments of the present invention.

[0053] FIG. 33 is a diagram illustrating scaling of motion information.

[0054] FIG. 34 is a diagram illustrating scaling of motion information in accordance with an embodiment of the present invention.

[0055] FIG. 35-39 are diagrams illustrating a method for updating and sharing a candidate list in accordance with some embodiments of the present invention.

[0056] FIG. 40 is a diagram illustrating a candidate list composed of one UPU and a candidate list composed of multiple UPUs in accordance with an embodiment of the present invention.

[0057] FIG. 41 is a flowchart illustrating a decoding method in accordance with an embodiment of the present invention.

[0058] FIG. 42 is a flowchart illustrating an encoding method in accordance with an embodiment of the present invention.

[0059] FIG. 43 and 44 are diagrams illustrating MPM candidates.

[0060] FIG. 45 is a diagram illustrating an example of intra-prediction mode selection.

[0061] FIG. 46 is a diagram illustrating a case where the intra-prediction mode of the current block is not included in the MPM list.

[0062] FIG. 47 is a diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[0063] FIG. 48 is another diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[0064] FIG. 49 illustrates a method illustrating a count-based list of intra-prediction modes in accordance with an embodiment of the present invention.

[0065] FIG. 50 is a diagram illustrating a method for configuring a cost-based intra-prediction mode list in accordance with an embodiment of the present invention.

[0066] FIG. 51 and 52 are diagrams illustrating a method for changing an intra-prediction mode list index in accordance with some embodiments of the present invention.

[0067] FIG. 53 and 54 are diagrams illustrating an area configuring an intra-prediction mode list in accordance with some embodiments of the present invention.

[0068] FIG. 55 is a diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[0069] FIG. 56-58 are diagrams illustrating a method for allocating an intra-prediction mode index in accordance with some embodiments of the present invention.

[0070] FIG. 59 is a diagram illustrating a method for configuring a plurality of intra-prediction mode lists in accordance with an embodiment of the present invention.

[0071] FIG. 60 and 61 are diagrams illustrating a method for configuring a plurality of intra-prediction mode lists in consideration of a prediction mode direction in accordance with an embodiment of the present invention.

[0072] FIG. 62 is a diagram illustrating the structure of an encoder and decoder for intra-prediction list completion according to an embodiment of the present invention.

[0073] FIG. 63 is a diagram illustrating a method for expanding a list of intra-prediction modes in accordance with an embodiment of the present invention.

[0074] FIG. 64 is a diagram illustrating a method for changing the substitution order taking into account the intra-prediction mode of a neighboring block in accordance with an embodiment of the present invention.

[0075] FIG. 65 is a diagram illustrating a method of applying a list of intra-prediction modes to an MPM mode in accordance with one embodiment of the present invention. Fig. 66 is a diagram for describing a method of applying an intra-prediction mode list to a non-MPM mode by an encoder in accordance with an embodiment of the present invention.

[0076] FIG. 66 is a diagram illustrating a method of applying an intra-prediction mode list to a non-MPM mode by an encoder in accordance with an embodiment of the present invention.

[0077] FIG. 67 is a diagram illustrating a method of applying an intra-prediction mode list to a non-MPM mode by a decoder according to an embodiment of the present invention.

[0078] FIG. 68 is a diagram illustrating an image encoding/decoding method using an intra-prediction mode list according to an embodiment of the present invention.

[0079] FIG. 69 is a diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[0080] FIG. 70 and 71 are diagrams illustrating an encoding/decoding method in accordance with some embodiments of the present invention.

[0081] FIG. 72 is a diagram illustrating a method for initializing an intra-prediction list in accordance with an embodiment of the present invention.

[0082] FIG. 73 is a diagram illustrating the syntax of an intra-prediction mode list according to an embodiment of the present invention.

Best mode

[0083] Many modifications can be made to the present invention, and there are various embodiments of the present invention, examples of which will be presented below and described in detail with reference to the drawings. However, the present invention is not limited to them, although the exemplary embodiments may be construed as including all modifications, equivalents or substitutions in the technical principle and technical scope of the present invention. Like reference numerals refer to the same or similar functions in different aspects. In drawings, shapes and sizes of elements may be exaggerated for clarity. In the following detailed description of the present invention, reference is made to the accompanying drawings, which show, by way of illustration, specific embodiments in which the present invention may be implemented. These embodiments are described in sufficient detail to enable those skilled in the art to implement the present disclosure. It should be understood that the numerous embodiments of the present disclosure, although different, are not necessarily mutually exclusive. For example, specific features, structures, and characteristics described herein in connection with one embodiment may be implemented in another embodiment without departing from the spirit and scope of the present disclosure. Moreover, it should be understood that the location or arrangement of individual elements in each disclosed embodiment may be modified without departing from the spirit and scope of the present disclosure. Therefore, the following detailed description is not to be construed in a limiting sense, and the scope of the present disclosure is determined only by the appended claims, construed as appropriate, together with the full range of equivalents of what is claimed in the claims.

[0084] The terms 'first', 'second', etc. used in the description. may be used to describe various components, but the components should not be implied to be limited by these terms. These terms are used only to distinguish one component from other components. For example, the 'first' component may be referred to as the 'second' component without departing from the scope of the present invention, and the 'second' component may also likewise be referred to as the 'first' component. The term 'and/or' includes a combination of a plurality of elements or any one of a plurality of elements.

[0085] It should be understood that when an element is simply referred to as being 'connected to' or 'associated with' another element, without being 'directly connected to' or 'directly associated with' the other element in the present specification, it may be 'directly connected to' ' or 'directly associated with' with another element or to be connected or associated with another element, having another element intermediate between them. On the contrary, it should be understood that when an element is referred to as “directly connected” or “directly connected” to another element, there are no intermediate elements.

[0086] In addition, the component parts shown in the embodiments of the present invention are shown independently to represent characteristic functions different from each other. Thus, this does not mean that each component part is contained in a separate hardware or software building block. That is, each component includes each of the numbered components for convenience. Thus, at least two component parts of each component part may be combined to form one component part, or one component part may be divided into multiple component parts to perform each function. An embodiment where each component part is combined and an embodiment where one component part is separated are also included within the scope of the present invention unless this leads to a departure from the spirit of the present invention.

[0087] The terms used in this description are used simply to describe specific embodiments and are not intended to limit the present invention. An expression used in the singular includes an expression in the plural unless it has a clearly different meaning in the context. As used herein, it should be understood that terms such as “including,” “having,” etc. are intended to indicate the presence of the features, quantities, steps, acts, elements, parts, or combinations thereof disclosed in the specification shall not exclude the possibility that one or more other features, quantities, steps, actions, elements, parts or combinations thereof may exist or be added. In other words, when a particular element is referred to as “included,” elements other than the corresponding element are not excluded, but additional elements may be included within embodiments of the present invention or the scope of the present invention.

[0088] In addition, some of the components may not be necessary components that perform the essential functions of the present invention, but are specific components that merely enhance its performance. The present invention can be implemented by including only the components necessary to implement the essence of the present invention, excluding components used to improve performance. A structure including only the necessary components, excluding specific components used only to improve performance, is also included within the scope of the present invention.

[0089] Next, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In describing exemplary embodiments of the present invention, well-known functions or structures will not be described in detail since they might unnecessarily complicate the understanding of the present invention. Like components in the drawings are designated by the same reference numerals, and repeated description of the same elements will be omitted.

[0090] Further, the image may refer to a picture configuring the video, or may refer to the video itself. For example, “encoding or decoding, or both, of a picture” may mean “encoding or decoding, or both, of a moving picture” and may mean “encoding or decoding, or both of them, one image among the pictures of a moving picture.”

[0091] Further, the terms “moving picture” and “video” may be used to have the same meaning and may be used interchangeably.

[0092] Further, the target image may be an encoding target image, which is an encoding target, and/or a decoding target image, which is a decoding target. Also, the target image may be an input image input to the encoding device and an input image input to the decoding device. Here, the target image can have the same value as the current image.

[0093] Further, the terms “image”, “picture”, “frame” and “screen” can be used to have the same meaning and replace each other.

[0094] Further, the target block may be an encoding target block, which is an encoding target, and/or a decoding target block, which is a decoding target. Also, the target block may be a current block that is the target of the current encoding and/or decoding. For example, the terms “target block” and “current block” can be used to have the same meaning and replace each other.

[0095] Further, the terms “block” and “unit” can be used to have the same meaning and replace each other. Or a “block” may represent a specific unit.

[0096] Further, the terms “region” and “segment” can be used interchangeably.

[0097] Further, the specific signal may be a signal representing a specific block. For example, the source signal may be a signal representing a target block. The prediction signal may be a signal representing a prediction block. The residual signal may be a signal representing a residual block.

[0098] In embodiments, each of specific information, data, flag, index, element and attribute, etc. may matter. An information, data, flag, index, element, and attribute value of “0” may represent a logical “false” or the first predefined value. In other words, the value "0", "false", Boolean "false", and the first predefined value can be used interchangeably. An information, data, flag, index, element, and attribute value of “1” can represent a logical “true” or a second predefined value. In other words, the value "1", "true", logical "true", and the second predefined value can be used interchangeably.

[0099] When a variable i or j is used to represent a column, row, or index, the value of i may be an integer equal to or greater than 0, or equal to or greater than 1. That is, column, row, index, etc. . can be counted from 0 or can be counted from 1.

[00100]

[00101] Description of terms

[00102] Encoder: means a device that performs encoding. That is, it means an encoding device.

[00103] Decoder: means a device that performs decoding. That is, it means a decoding device.

[00104] Block: is an M×N sample array. Here, M and N can mean positive integers, and block can mean a two-dimensional shaped sampling array. A block can refer to a unit. The current block can mean an encoding target block, which becomes a target when encoding, or a decoding target block, which becomes a target when decoding. In addition, the current block may be at least one of an encoding block, a prediction block, a residual block, and a transform block.

[00105] Sample: represents the basic unit that forms a block. It can be expressed as a value from 0 to 2Bd-1 according to the bit depth (Bd). In the present invention, a sample can be used to designate a pixel. That is, a sample, an image element, a pixel can have the same meaning.

[00106] Unit: May refer to a unit of encoding and decoding. When encoding and decoding an image, a unit may be an area generated by dividing a single image. In addition, a unit can mean a subdivided unit, where one image is divided into subdivided units during encoding or decoding. That is, the image can be divided into many units. When encoding and decoding an image, a predefined process may be performed for each unit. One unit can be divided into sub-units that are smaller than the unit size. Depending on the functions, a unit may mean a block, a macroblock, a coding tree unit, a coding tree block, a coding unit, a coding block, a prediction unit, a prediction block, a residual unit, a residual block, a transform unit, a transform block, etc. In addition, to distinguish a unit from a block, the unit may include a luma component block, a chroma component block associated with the luma component block, and a syntax element of each color component block. The unit may have various sizes and shapes, in particular, the shape of the unit may be a two-dimensional geometric shape such as a square shape, a rectangular shape, a trapezoidal shape, a triangular shape, a pentagonal shape, etc. In addition, the unit information may include at least one of a unit type indicating a coding unit, a prediction unit, a transformation unit, etc., and a unit size, a unit depth, a unit encoding and decoding sequence, etc.

[00107] Coding tree unit: configured with one coding tree block of the Y luma component and two coding tree blocks related to the Cb and Cr chrominance components. It may also mean that blocks and the syntax element of each block are included. Each coding tree unit may be subdivided using at least one of a quadtree splitting method, a binary tree splitting method, and a ternary tree splitting method to configure a lower unit such as a coding unit, a prediction unit, a transformation unit, etc. This can be used as a term for a sample block that becomes a processing unit when encoding/decoding an image as an input image. Here, quadtree can mean quaternary tree.

[00108] When the encoding block size is within a predetermined range, division is possible using only quadtree partitioning. Here, a predetermined range may be defined as at least one of a maximum size and a minimum encoding block size in which division is possible using only quadtree partitioning. Information indicating the maximum/minimum size of a coding block in which quadtree partitioning is enabled may be signaled by the bitstream, and the information may be signaled in at least one sequence unit, picture parameter, tile group, or slice. Alternatively, the maximum/minimum encoding block size may be a fixed size predefined in the encoder/decoder. For example, when the encoding block size corresponds to 256×256 to 64×64, division is only possible using quadtree partitioning. Alternatively, when the encoding block size is larger than the maximum transform block size, partitioning is only possible using quadtree partitioning. Here, the block to be divided may be at least one of an encoding block and a transform block. In this case, the information indicating splitting of the coded block (eg, split_flag) may be a flag indicating whether or not to perform quadtree splitting. When the encoding block size falls within a predefined range, partitioning is only possible using binary tree or ternary tree partitioning. In this case, the above description of quadtree partitioning can be applied to binary tree partitioning or ternary tree partitioning in the same way.

[00109] Coding tree block: may be used as a term to refer to any one of a Y coding tree block, a Cb coding tree block, and a Cr coding tree block.

[00110] Adjacent block: may mean a block adjacent to the current block. A block adjacent to the current block may mean a block that is in contact with the boundary of the current block, or a block positioned within a predetermined distance from the current block. Adjacent block can mean a block adjacent to the top of the current block. Here, a block adjacent to the top of the current block may mean a block vertically adjacent to an adjacent block that is horizontally adjacent to the current block, or a block horizontally adjacent to an adjacent block that is vertically adjacent to the current block.

[00111] Recovered neighboring block: may mean a neighboring block adjacent to the current block and which has already been spatially/temporally encoded or decoded. Here, a restored neighbor block may mean a restored neighbor unit. The reconstructed spatial neighbor block may be a block within the current picture and has already been recovered through encoding or decoding or both. The reconstructed temporary neighboring block is a block in the corresponding position as the current block of the current picture within the reference image or its neighboring block.

[00112] Unit depth: may refer to the degree to which a unit is subdivided. In a tree structure, the highest node (root node) may correspond to the first unit that is not split. Also, the highest node may have the smallest depth value. In this case, the highest node may have a depth of level 0. A node having a depth of level 1 may represent a unit generated by splitting the first unit once. A node having a depth of level 2 can represent a unit generated by splitting the first unit twice. A node having level depth n can represent a unit generated by splitting the first unit n times. A leaf node can be the lowest node and a node that cannot split further. The depth of the leaf node can correspond to the maximum level. For example, the predefined maximum level value could be 3. The root node depth could be the lowest, and the leaf node depth could be the deepest. Additionally, when a unit is expressed as a tree structure, the level at which the unit is present can indicate the depth of the unit.

[00113] Bitstream: may mean a bitstream including image encoding information.

[00114] Parameter set: corresponds to the header information among the configuration in the bitstream. At least one of a video parameter set, a sequence parameter set, a picture parameter set, and an adaptation parameter set may be included in the parameter set. In addition, the set of parameters may include a slice header, a tile group header, and tile header information. The term “tile group” means a group of tiles and has the same meaning as a slice.

[00115] Parsing: can mean determining the meaning of a syntactic element by performing entropy decoding, or can mean entropy decoding itself.

[00116] Symbol: may represent at least one of a syntax element, an encoding parameter, and a conversion factor value of the target encoding/decoding unit. In addition, the symbol can indicate the target of entropy encoding or the result of entropy decoding.

[00117] Prediction mode: may be information indicating an intra-prediction encoded/decoded mode or an inter-prediction encoded/decoded mode.

[00118] Prediction unit: may mean a basic unit when performing prediction such as inter-prediction, intra-prediction, inter-(mutual) compensation, intra-(internal) compensation and motion compensation. One prediction unit may be split into multiple splits having smaller dimensions, or may be split into multiple smaller prediction units. A plurality of partitions may represent the basic unit when performing prediction or compensation. A partition that is generated by dividing a prediction unit may also be a prediction unit.

[00119] Prediction unit split: may refer to a shape obtained by splitting a prediction unit.

[00120] A reference picture list may refer to a list including one or more reference pictures used for inter-prediction or motion compensation. There are several types of reference picture lists used, including LC (List Concatenated), L0 (List 0), L1 (List 1), L2 (List 2), L3 (List 3).

[00121] The inter-prediction indicator may refer to the inter-prediction direction (unidirectional prediction, bidirectional prediction, etc.) of the current block. Alternatively, it may refer to the number of reference pictures used to generate the prediction block of the current block. Alternatively, it may refer to the number of prediction blocks used while performing inter-prediction or motion compensation on the current block.

[00122] The prediction list use flag indicates whether a prediction block is generated using at least one reference picture in a particular reference picture list. An inter-prediction indicator may be output using a prediction list use flag, and conversely, a prediction list use flag may be output using an inter-prediction indicator. For example, when the prediction list use flag has a first value of zero (0), it means that a reference picture in the reference picture list is not used to generate a prediction block. On the other hand, when the prediction list use flag has a second value of one (1), it means that the reference picture list is used to generate the prediction block.

[00123] The reference picture index may refer to an index indicating a specific reference picture in the list of reference pictures.

[00124] A reference picture may refer to a reference picture that is referenced by a particular block for the purposes of inter-prediction or motion compensation of the particular block. Alternatively, the reference picture may be a picture including a reference block referenced by the current block for inter-prediction or motion compensation. Further, the terms “reference picture” and “reference image” have the same meaning and can be used interchangeably.

[00125] The motion vector may be a two-dimensional vector used for inter-prediction or motion compensation. The motion vector may indicate an offset between a target encoding/decoding block and a reference block. For example, (mvX, mvY) can represent a motion vector. Here, mvX may represent a horizontal component, and mvY may represent a vertical component.

[00126] The search range may be a two-dimensional region in which a search is performed to extract a motion vector during inter-prediction. For example, the size of the search range may be M×N. Here, M and N are both integers.

[00127] The motion vector candidate may refer to a prediction candidate block or a motion vector of a prediction candidate block when predicting a motion vector. In addition, the motion vector candidate may be included in the motion vector candidate list.

[00128] A motion vector candidate list may mean a list composed of one or more motion vector candidates.

[00129] The motion vector candidate index may mean an index indicating the motion vector candidate in the motion vector candidate list. Alternatively, it may be a motion vector predictor index.

[00130] Motion information may mean information including at least one of a motion vector, a reference picture index, an inter-prediction indicator, a prediction list use flag, reference picture list information, a reference picture, a motion vector candidate, an index motion vector candidate, fusion candidate and fusion index.

[00131] A merge candidate list may refer to a list composed of one or more merge candidates.

[00132] A merge candidate may mean a spatial merge candidate, a temporal merge candidate, a pooled merge candidate, a pooled bi-predictive merge candidate, or a null merge candidate. The fusion candidate may include motion information such as an inter-prediction indicator, a reference picture index for each list, a motion vector, a prediction list usage flag, and an inter-prediction indicator.

[00133] A merge index may mean a pointer indicating a merge candidate in a list of merge candidates. Alternatively, the merge index may indicate the block from which the merge candidate was derived among reconstructed blocks spatially/temporally adjacent to the current block. Alternatively, the merger index may indicate at least one piece of merger candidate traffic information.

[00134] Transformation unit: may mean a basic unit when performing encoding/decoding, such as transforming, inverse transforming, quantizing, dequantizing, encoding/decoding of residual signal transform coefficients. One conversion unit can be broken down into many lower level conversion units of smaller size. Here, the transform/inverse transform may comprise at least one of a first transform/first inverse transform and a second transform/second inverse transform.

[00135] Scaling: may refer to the process of multiplying the quantized level by a factor. The conversion factor can be generated by scaling the quantized level. Scaling may also be referred to as dequantization.

[00136] Quantization parameter: may mean a value used when generating a quantized level using a quantization conversion factor. The quantization parameter can also mean the value used in generating the conversion factor by scaling the quantized level during dequantization. The quantization parameter may be a value mapped to a quantization step size.

[00137] Delta quantization parameter: may mean the difference value between the predicted quantization parameter and the quantization parameter of the target encoding/decoding unit.

[00138] Scan: may refer to a method of ordering coefficients within a unit, block, or matrix. For example, the transition from a two-dimensional coefficient matrix to a one-dimensional matrix may be referred to as a scan, and the transition from a one-dimensional coefficient matrix to a two-dimensional matrix may be referred to as a scan or reverse scan.

[00139] Transform coefficient: may refer to the coefficient value generated after performing the transformation in the encoder. It may mean a coefficient value generated after performing at least one of entropy decoding and dequantization in a decoder. The quantized level obtained by quantizing the transform coefficient or residual signal, or the quantized transform coefficient level may also be covered by the semantic value of the transform coefficient.

[00140] Quantized level: may mean the value generated by quantizing the transform coefficient or residual signal in the encoder. Alternatively, the quantized level may mean a value that is a dequantization target for performing dequantization at the decoder. Likewise, the quantized transform coefficient level, which is the result of the transform and quantization, can also be captured by the semantic value of the quantized level.

[00141] Non-zero transform coefficient: may mean a transform coefficient having a value other than zero, or a transform coefficient level or quantized level having a value other than zero.

[00142] Quantization matrix: may refer to a matrix used in a quantization process or a dequantization process performed to improve the subjective or objective quality of an image. The quantization matrix can also be referred to as a scaling list.

[00143] Quantization matrix coefficient: may refer to each element in the quantization matrix. The quantization matrix coefficient may also be referred to as the matrix coefficient.

[00144] Default matrix: may mean a predefined quantization matrix predefined in the encoder or decoder.

[00145] Non-default matrix: may mean a quantization matrix that is not predefined in the encoder or decoder, but is signaled by the user.

[00146] Statistical value: the statistical value for at least one of a variable, encoding parameter, constant value, etc. that has a particular value calculated, may be one or more of an average value, a total value, a weighted average value, a weighted value total value, minimum value, maximum value, most frequent value, median value, interpolated value of the corresponding specific value.

[00147] FIG. 1 is a block diagram showing the configuration of an encoding device in accordance with an embodiment to which the present invention is applicable.

[00148] The encoding device 100 may be an encoder, a video encoding device, or an image encoding device. The video may include at least one image. The encoding device 100 may sequentially encode at least one image.

[00149] With reference to FIG. 1, the encoding device 100 may include a motion prediction module 111, a motion compensation module 112, an intra-prediction module 120, a switch 115, a subtractor 125, a transform module 130, a quantization module 140, an entropy encoding module 150, a dequantization module 160, a module 170 an inverse transform, an adder 175, a filtering module 180, and a reference picture buffer 190.

[00150] The encoding device 100 may encode the input image using intra-mode or inter-mode or both. In addition, the encoding device 100 can generate a bitstream including encoded information by encoding an input image, and output the generated bitstream. The generated bit stream may be stored in a computer-readable recording medium or may be streamed over a wired/wireless transmission medium. When the intra mode is used as the prediction mode, the switch 115 may be switched to the intra mode. Alternatively, when inter mode is used as a prediction mode, switch 115 may be switched to inter mode. Here, intra-mode may mean an intra-prediction mode, and inter-mode may mean an inter-prediction mode. The encoding device 100 may generate a prediction block for an input block of an input image. In addition, the encoding device 100 may encode the residual block using the remainder of the input block and the prediction block after generating the prediction block. The input image may be called the current image, which represents the current encoding target. The input block may be called the current block, which represents the current encoding target, or the encoding target block.

[00151] When the prediction mode is intra-mode, intra-prediction module 120 may use a sample of a block that has already been encoded/decoded and is adjacent to the current block as a reference sample. The intra-prediction module 120 may perform spatial prediction on the current block using a reference sample, or generate prediction samples of the input block by performing spatial prediction. Here, internal (intra) prediction can mean intra-prediction,

[00152] When the prediction mode is inter-mode, the motion prediction unit 111 can extract the region that best matches the input block from the reference image when performing motion prediction, and output a motion vector using the extracted region. In this case, the search area can be used as the mentioned area. The reference picture may be stored in the reference picture buffer 190. Here, when encoding/decoding for a reference picture is completed, it can be stored in the reference picture buffer 190.

[00153] The motion compensation unit 112 can generate a prediction block by performing motion compensation on the current block using a motion vector. Here, inter-prediction may mean inter-prediction or motion compensation.

[00154] When the motion vector value is not an integer, the motion prediction unit 111 and the motion compensation unit 112 can generate a prediction block by applying an interpolation filter to a partial area of the reference picture. To perform inter-picture prediction or motion compensation on a coding unit, it may be determined which mode among the skip mode, merge mode, advanced motion vector prediction (AMVP) mode, and current picture reference mode is used for motion prediction and motion compensation of the prediction unit included in the corresponding coding unit. Then, inter-picture prediction or motion compensation may be performed differently depending on the specific mode.

[00155] Subtractor 125 may generate a residual block using the difference of the input block and the prediction block. The residual block may be called a residual signal. The residual signal may mean the difference between the original signal and the prediction signal. Moreover, the residual signal may be a signal generated by transforming or quantizing, or transforming and quantizing the difference between the original signal and the prediction signal. The residual block may be a residual block unit signal.

[00156] The transformation unit 130 can generate a transformation coefficient by performing transformation on a residual block and output the generated transformation coefficient. Here, the transformation coefficient may be a coefficient value generated by performing transformation of the residual block. When the skip conversion mode is applied, the conversion unit 130 may skip the conversion of the residual block.

[00157] The quantized level can be generated by applying quantization to the transform coefficient or to the residual signal. Further, the quantized level may also be referred to as a conversion coefficient in embodiments.

[00158] The quantization unit 140 can generate a quantized level by quantizing a transform coefficient or a residual signal according to a parameter, and output the generated quantized level. Here, the quantization unit 140 may quantize the transform coefficient using a quantization matrix.

[00159] The entropy encoding unit 150 can generate a bit stream by performing entropy encoding according to a probability distribution on the values calculated by the quantization unit 140 or on the encoding parameter values calculated when performing encoding, and output the generated bit stream. The entropy encoding unit 150 can perform entropy encoding of image sampling information and image decoding information. For example, the image decoding information may include a syntax element.

[00160] When entropy encoding is used, symbols are represented such that fewer bits are assigned to a symbol having a high probability of being generated and a larger number of bits are assigned to a symbol having a low probability of being generated, and thus the size of the bit stream for the symbols to be encoded , may decrease. The entropy coding unit 150 may use a coding method for entropy coding such as exponential Golomb coding, context adaptive variable length coding (CAVLC), context adaptive binary arithmetic coding (CABAC), etc. For example, entropy encoding module 150 may perform entropy encoding using a codebook/variable length code (VLC). In addition, the entropy encoding module 150 can output the binarization method of the target symbol and the probability model of the target symbol/bin, and perform arithmetic encoding using the derived binarization method and context model.

[00161] To encode the transform coefficient level (quantized level), the entropy encoding module 150 can transform the coefficient in the form of a two-dimensional block into a one-dimensional vector form using a transform coefficient scanning method.

[00162] The encoding parameter may include information (flag, index, etc.) such as a syntax element that is encoded in the encoder and signaled to the decoder, and information output when encoding or decoding is performed. An encoding parameter may refer to information required when encoding or decoding an image. For example, at least one value or form of a combination of unit/block size, unit/block depth, unit/block partitioning information, unit/block shape, unit/block partition structure, whether to partition in the form of a quadtree, whether to partition in the form binary tree, the direction of partitioning of the binary tree form (horizontal direction or vertical direction), the partitioning shape of the binary tree form (symmetric partitioning or asymmetric partitioning), whether the current coding unit is partitioned by partitioning the ternary tree, the direction (horizontal or vertical direction) of partitioning the ternary tree , the type (symmetric or asymmetric type) of the ternary tree partition, whether the current encoding unit is partitioned by a multi-type tree partition, the direction (horizontal or vertical direction) of the multi-type tree partition, the type (symmetric or asymmetric type) of the multi-type tree partition, and the structure of the tree ( binary tree or ternary tree) tree partitions of multiple types, prediction mode (intra-prediction or inter-prediction), luma intra-prediction mode/direction, chrominance intra-prediction mode/direction, intra-partition information, inter-partition information, flag encoding block split flag, prediction block split flag, transform block split flag, reference sample filtering method, reference sample filter tap, reference sample filter coefficient, prediction block filtering method, prediction block filter tap, prediction block filter coefficient, prediction block boundary filtering method, prediction block boundary filter tap, prediction block boundary filter coefficient, intra-prediction mode, inter-prediction mode, motion information, motion vector, motion vector difference, reference picture index, inter-prediction angle, inter-prediction pointer, prediction list use flag , reference picture list, reference picture, motion vector predictor index, motion vector predictor candidate, motion vector candidate list, whether to use merge mode, merge index, merge candidate, merge candidate list, whether to use skip mode, interpolation filter type, tap interpolation filter, interpolation filter coefficient, motion vector size, motion vector representation precision, transformation type, transformation size, information whether the primary (first) transformation is used, information whether the secondary transformation is used, primary transformation index, secondary transformation index, information, represented whether residual signal, coded block pattern, coded block flag (CBF), quantization parameter, residual quantization parameter, quantization matrix, whether to apply an in-loop filter, in-loop filter coefficient, in-loop filter tap, in-loop filter shape, whether to apply a deblocking filter, deblocking filter coefficient, deblocking filter tap, deblocking filter strength, deblocking filter shape, whether to apply adaptive sample offset, adaptive sample offset value, adaptive sample offset category, adaptive sample offset type, whether to apply adaptive loop filter, coefficient adaptive loop filter, adaptive loop filter tap, adaptive loop filter shape, binarization/inverse binarization method, context model determination method, context model update method, whether to perform regular mode, whether to perform bypass mode, context bin, bypass bin, significant flag coefficient, the last significant coefficient flag, the coded flag for the unit of the coefficient group, the position of the last significant coefficient, the flag for whether the coefficient value is greater than 1, the flag for whether the coefficient value is greater than 2, the flag for whether whether the coefficient value is greater than 3, residual coefficient value information, sign information, reconstructed luminance sample, reconstructed chroma sample, residual luminance sample, residual chrominance sample, luminance conversion coefficient, chrominance conversion coefficient, quantized luminance level, quantized chrominance level, method conversion coefficient level scanning, decoder side motion vector search area size, decoder side motion vector search area shape, decoder side motion vector search number of times, CTU size information, minimum block size information, maximum block size information, information maximum block depth information, minimum block depth information, image display/output sequence, slice identification information, slice type, slice partition information, tile identification information, tile type, tile partition information, tile group identification information, tile group type elements, tile group partitioning information, picture type, input sample bit depth, reconstructed sample bit depth, residual sample bit depth, conversion coefficient bit depth, quantized level bit depth and luma signal information or chrominance signal information may be included in the parameter coding.

[00163] Here, signaling a flag or index may mean that the corresponding flag or index is entropy encoded and included in the bit stream by the encoder, and may mean that the corresponding flag or index is entropy decoded from the bit stream by the decoder.

[00164] When the encoding device 100 performs inter-prediction encoding, the encoded current image can be used as a reference image for another image that is processed thereafter. Accordingly, the encoding apparatus 100 may recover or decode the encoded current picture, or store the recovered or decoded image as a reference picture in the reference picture buffer 190.

[00165] The quantized level may be dequantized in dequantization module 160, or may be dequantized in dequantization module 170. The dequantized and/or inversely transformed coefficient may be added to the prediction block via adder 175. By adding the dequantized and/or inversely transformed coefficient to the prediction block, a reconstructed block may be generated. Here, a dequantized and/or de-converted coefficient may mean a coefficient on which at least one of dequantization and de-conversion has been performed, and may denote a reconstructed residual block.

[00166] The recovered block may pass through filter module 180. Filtering module 180 may apply at least one of a deblocking filter, an adaptive sample offset (SAO), and an adaptive loop filter (ALF) to the reconstructed sample, the reconstructed block, or the reconstructed image. Filtration module 180 may be referred to as a loop filter.

[00167] A deblocking filter may remove block artifact generated at boundaries between blocks. To determine whether the deblocking filter should be applied, it may be determined whether the deblocking filter should be applied to the current block based on the samples included in the various rows or columns that are included in the block. When a deblocking filter is applied to a block, another filter may be applied according to the desired strength (level) of deblocking filtering.

[00168] In order to compensate for encoding error, an appropriate offset value can be added to the sample value using an adaptive sample offset. Adaptive sampling offset can adjust the offset of the deblocked image relative to the original image by a sampling unit. A method may be used of dividing the image samples into a predetermined number of regions, determining the region to which the offset is applied and applying the offset to the defined region, or a method of applying the offset taking into account edge information on each sample.

[00169] The adaptive edge filter may perform filtering based on a comparison result between the filtered reconstructed image and the original image. The samples included in the image can be divided into predefined groups, a filter can be defined to apply to each group, and differential filtering can be performed on each group. Information about whether ALF should be applied may be signaled by coding units (CUs), and the shape and ratio of the ALF to be applied to each block may vary.

[00170] The reconstructed block or reconstructed image, after passing through the filtering module 180, may be stored in the reference picture buffer 190. The reconstructed block processed by filtering module 180 may be part of a reference image. That is, the reference image is a reconstructed image composed of reconstructed blocks processed by the filtering unit 180. The stored reference image can be used later in inter-prediction or motion compensation.

[00171] FIG. 2 is a block diagram showing the configuration of a decoding apparatus according to an embodiment to which the present invention is applicable.

[00172] The decoding device 200 may be a decoder, a video decoding device, or an image decoding device.

[00173] With reference to FIG. 2, the decoding apparatus 200 may include an entropy decoding unit 210, a dequantization unit 220, an inverse transform unit 230, an intra-prediction unit 240, a motion compensation unit 250, an adder 225, a filtering unit 260, and a reference picture buffer 270.

[00174] The decoding device 200 may receive a bit stream output from the encoding device 100. The decoding apparatus 200 may receive a bit stream stored in a computer-readable recording medium, or may receive a bit stream transmitted by streaming via a wired/wireless transmission medium. The decoding apparatus 200 may decode the bit stream using intra mode or inter mode. In addition, the decoding apparatus 200 can generate a reconstructed image generated by decoding or a decoded image and output the reconstructed image or decoded image.

[00175] When the prediction mode used in decoding is intra mode, the switch may be switched to intra mode. Alternatively, when the prediction mode used in decoding is inter-mode, the switch may be switched to inter-mode.

[00176] The decoding apparatus 200 can obtain the reconstructed residual block by decoding the input bit stream and generate a prediction block. When the reconstructed residual block and the prediction block are received, the decoding apparatus 200 can generate a reconstructed block that becomes a decoding target by summing the reconstructed residual block with the prediction block. The decoding target block may be called the current block.

[00177] The entropy decoding unit 210 may generate symbols by entropy decoding a bitstream according to a probability distribution. The generated symbols may include a quantized level shape symbol. Here, the entropy decoding method may be a reverse process to the entropy encoding method described above.

[00178] In order to decode the transform coefficient level (quantized level), the entropy decoding unit 210 can transform the coefficient in the form of a one-dimensional vector into the form of a two-dimensional block using a transform coefficient scanning method.

[00179] The quantized level may be dequantized in dequantization module 220 or inverted in dequantization module 230. The quantized level may be the result of dequantization and/or deconversion and may be generated as a reconstructed residual block. Here, dequantization module 220 may apply a quantization matrix to the quantized level.

[00180] When the intra mode is used, the intra prediction module 240 can generate a prediction block by performing, for the current block, a spatial prediction that uses the sample value of a block adjacent to the decoding target block and that has already been decoded.

[00181] When inter-mode is used, motion compensation module 250 can generate a prediction block by performing, for the current block, motion compensation that uses a motion vector and a reference picture stored in the reference picture buffer 270.

[00182] Adder 225 may generate a reconstructed block by adding the reconstructed residual block to the prediction block. Filtering module 260 may apply at least one of a deblocking filter, an adaptive sample offset, and an adaptive contour filter to the reconstructed block or the reconstructed image. The filtering module 260 may output the reconstructed image. The reconstructed block or reconstructed image may be stored in reference picture buffer 270 and used when performing inter-prediction. The reconstructed block processed by filtering module 260 may be part of a reference image. That is, the reference image is a reconstructed image composed of reconstructed blocks processed by the filtering module 260. The saved reference image can be used later for inter-prediction or motion compensation.

[00183] FIG. 3 is a view schematically showing an image division structure in image encoding and decoding. Fig. 3 schematically shows an example of breaking one unit into many lower units.

[00184] In order to efficiently partition an image during encoding and decoding, a coding unit (CU) may be used. The coding unit can be used as a basic unit when encoding/decoding an image. In addition, the encoding unit can be used as a unit for distinguishing between an intra-prediction mode and an inter-prediction mode when encoding/decoding an image. The coding unit may be a basic unit used for prediction, transform, quantization, inverse transform, dequantization, or a transform coefficient encoding/decoding process.

[00185] With reference to FIG. 3, the image 300 is sequentially divided into a largest coding unit (LCU), and the LCU is defined as a division structure. Here, LCU can be used in the same sense as a coding tree unit (CTU). Splitting a unit may mean splitting the block associated with the unit. Unit depth information may be included in the block splitting information. The depth information may represent the number of times and/or the degree to which a unit is split. One unit may be broken down into multiple lower level units hierarchically associated with depth information based on a tree structure. In other words, a certain unit and a lower level unit generated by splitting a given unit can correspond to a node and a child node of a given node, respectively. Each divided lower unit may have depth information. The depth information may be information representing the size of a CU and may be stored in each CU. The depth of a unit represents the number of times and/or degrees of division of a unit. Therefore, the partitioning information of a lower-level unit may contain information about the size of the lower-level unit.

[00186] A partitioning structure may refer to the allocation of a coding unit (CU) within the LCU 310. Such allocation may be determined according to whether one CU should be partitioned into a plurality (a positive integer equal to or greater than 2, including 2, 4, 8, 16, etc.) CU. The horizontal size and vertical size of the CU generated by partitioning may respectively be half the horizontal size and vertical size of the CU before partitioning, or may suitably have dimensions smaller than the horizontal size and vertical size before partitioning, according to the number of times of partitioning. A CU can be recursively split into multiple CUs. By recursively partitioning, at least one of the height and width of the CU after the partition can be reduced compared to at least one of the height and width of the CU before the partition. CU partitioning can be performed recursively to a predefined depth or predefined size. For example, the LCU depth may be 0, and the smallest coding unit (SCU) depth may be a predefined maximum depth. Here, the LCU may be a coding unit having a maximum coding unit size, and the SCU may be a coding unit having a minimum coding unit size, as described above. Partitioning starts from LCU 310, the CU depth increases by 1 as the horizontal size and/or vertical size of the CU is reduced by partitioning. For example, for each depth, a CU that is not partitioned may have a size of 2N×2N. Also, in the case of a CU that is partitioned, a 2N×2N size CU may be split into four N×N size CUs. The size N can be reduced by half when the depth increases by 1.

[00187] In addition, information about whether the CU should be split may be provided using the CU split information. The partition information may be 1-bit information. All CUs except SCU may include partitioning information. For example, when the value of the partition information is equal to the first value, the CU cannot be partitioned; when the value of the partition information is equal to the second value, the CU can be partitioned.

[00188] With reference to FIG. 3, an LCU having a depth of 0 may be a 64x64 block. 0 may be the minimum depth. An SCU having a depth of 3 may be an 8x8 block. 3 may be the maximum depth. CUs in the form of 32x32 blocks and 16x16 blocks can be represented as depth 1 and depth 2, respectively.

[00189] For example, when one coding unit is divided into four coding units, the horizontal size and vertical size of the four divided coding units may be equal to half the size of the horizontal and vertical size of the CU before the division. In one embodiment, when a coding unit having a size of 32×32 is divided into four coding units, each of the four divided coding units may have a size of 16×16. When one coding unit is split into four coding units, it can be said that the coding unit can be split into a quadtree shape.

[00190] For example, when one encoding unit is split into two sub-encoding units, the horizontal or vertical size (width or height) of each of the two sub-encoding units may be half the horizontal or vertical size of the original encoding unit. For example, when a coding unit having a size of 32×32 is vertically divided into two sub-coding units, each of the two sub-coding units may have a size of 16×32. For example, when a coding unit having a size of 8x32 is horizontally divided into two sub-coding units, each of the two sub-coding units may have a size of 8x16. When one coding unit is split into two sub-coding units, the coding unit can be said to be binary split or split through a binary tree split structure.

[00191] For example, when one encoding unit is divided into three sub-encoding units, the horizontal or vertical size of the encoding unit may be divided in the ratio 1:2:1, thereby forming three sub-encoding units whose horizontal or vertical dimensions are in the ratio 1:2:1. For example, when a coding unit having a size of 16×32 is horizontally divided into three sub-coding units, the three sub-coding units may have sizes of 16×8, 16×16 and 16×8, respectively, in order from the highest to the highest lower sub-unit coding. For example, when a coding unit having a size of 32×32 is vertically divided into three sub-coding units, the three sub-coding units may have sizes of 8×32, 16×32 and 8×32, respectively, in order from left to right sub-units. coding units. When one coding unit is split into three sub-coding units, the coding unit can be said to be ternary split or split through a ternary tree split structure.

[00192] In FIG. 3, a coding tree unit (CTU) 320 is an example of a CTU to which a quadtree partitioning structure, a binary tree partitioning structure, and a ternary tree partitioning structure are applied.

[00193] As described above, in order to split the CTU, at least one of a quadtree splitting structure, a binary tree splitting structure, and a ternary tree splitting structure can be applied. Different tree partitioning structures can be applied sequentially to CTUs, according to a predefined order of priority. For example, a quadtree partitioning structure may advantageously apply to a CTU. A coding unit that can no longer be split using the quadtree splitting structure may correspond to a quadtree leaf node. The encoding unit corresponding to a leaf node of the quadtree may serve as the root node of the binary and/or ternary tree partition structure. That is, the encoding unit corresponding to a quadtree leaf node may be further partitioned according to a binary tree partitioning structure or a ternary tree partitioning structure, or may not be partitioned any more. Therefore, by prohibiting a coding block that results from binary tree splitting or ternary tree splitting of a coding unit corresponding to a quadtree leaf node from undergoing further quadtree splitting, block splitting and/or signaling of splitting information can be effectively performed.

[00194] The fact that a coding unit corresponding to a quadtree node is partitioned may be signaled using quad-partition information. The quad-split information having a first value (eg, “1”) may indicate that the current encoding unit is split according to a quad-tree split structure. Quad-split information having a second value (eg, “0”) may indicate that the current encoding unit is not split according to a quad-tree split structure. The quad partition information may be a flag having a predetermined length (eg, one bit).

[00195] There may be no priority between binary tree splitting and ternary tree splitting. That is, the encoding unit corresponding to a leaf node of the quadtree may further be subject to arbitrary partitioning from a binary tree partition and a ternary tree partition. In addition, the encoding unit generated by binary tree splitting or ternary tree splitting may be subject to further binary tree splitting or further ternary tree splitting, or may not be further split.

[00196] A tree structure in which there is no priority among binary tree partitioning and ternary tree partitioning is referred to as a multi-type tree structure. The encoding unit corresponding to a leaf node of a quadtree can serve as the root node of a tree of multiple types. Whether or not to split an encoding unit that corresponds to a node of a multi-type tree may be signaled using at least one of multi-type tree splitting indication information, splitting direction information, and splitting tree information. To partition an encoding unit corresponding to a node of a multi-type tree, the multi-type tree partitioning indication information, the partitioning direction, and the partition tree information may be signaled sequentially.

[00197] The multi-type tree splitting indication information having a first value (eg, “1”) may indicate that the current encoding unit should undergo multi-type tree splitting. The multitype tree splitting indication information having a second value (eg, “0”) may indicate that the current encoding unit should not undergo multitype tree splitting.

[00198] When an encoding unit corresponding to a node of a multi-type tree is further partitioned according to a partitioning structure of the multi-type tree, the encoding unit may include partitioning direction information. The split direction information may indicate in which direction the current encoding unit should be split to split a tree of multiple types. The split direction information having a first value (eg, “1”) may indicate that the current coding unit should be split vertically. The split direction information having a second value (eg, “0”) may indicate that the current coding unit should be split horizontally.

[00199] When an encoding unit corresponding to a multi-type tree node is further partitioned according to a multi-type tree partitioning structure, the current encoding unit may include partition tree information. The partitioning tree information may indicate a partitioning tree structure that should be used to partition a tree node of the plurality of types. The split tree information having a first value (eg, “1”) may indicate that the current encoding unit should be split according to the binary tree split structure. The split tree information having a second value (eg, “0”) may indicate that the current coding unit should be split according to a ternary tree split structure.

[00200] The partition indication information, the partition tree information, and the partition direction information may each be a flag having a predetermined length (eg, one bit).

[00201] At least any one of the quad-tree partition indication information, the multi-type tree partition indication information, the partition direction information, and the partition tree information may be entropy encoded/decoded. For entropy encoding/decoding of these types of information, information about a neighboring coding unit adjacent to the current coding unit can be used. For example, there is a high probability that the split type (split or unsplit, split tree and/or split direction) of the left adjacent coding unit and/or top neighboring coding unit of the current coding unit is similar to that of the current coding unit. Therefore, context information for entropy encoding/decoding information about the current encoding unit can be inferred from information about neighboring encoding units. The neighbor encoding unit information may include at least any one of quad-partitioning information, multi-type tree partitioning indication information, partitioning direction information, and partitioning tree information.

[00202] As another example, from binary tree splitting and ternary tree splitting, binary tree splitting may preferably be performed. That is, the current encoding unit may be mainly subjected to binary tree splitting, and then the encoding unit corresponding to a leaf node of the binary tree may be set as a root node for ternary tree splitting. In this case, neither the quadtree split nor the binary tree split may be performed on the coding unit corresponding to the ternary tree node.

[00203] An encoding unit that cannot be partitioned according to a quadtree partition structure, a binary tree partition structure, and/or a ternary tree partition structure becomes a basic unit for encoding, prediction, and/or transformation. That is, the coding unit cannot be further broken down for prediction and/or transformation. Therefore, the partitioning structure information and the partitioning information used to partition the encoding unit into prediction units and/or transformation units cannot be represented in the bit stream.

[00204] However, when the size of the coding unit (i.e., the base unit for partitioning) is larger than the size of the maximum transform block, the coding unit may be recursively divided until the size of the coding unit is reduced to be equal to or less than than the size of the maximum conversion block. For example, when the coding unit size is 64×64 and when the maximum transform block size is 32×32, the coding unit may be divided into four 32×32 transform blocks. For example, when the coding unit size is 32×64 and the maximum transform block size is 32×32, the coding unit may be divided into two 32×32 transform blocks. In this case, the partition of the coding unit for transformation is not signaled separately and can be determined by comparison between the horizontal or vertical size of the coding unit and the horizontal or vertical size of the maximum transformation block. For example, when the horizontal size (width) of a coding unit is larger than the horizontal size (width) of a maximum transform block, the coding unit may be vertically divided in half. For example, when the vertical size (length) of a coding unit is larger than the vertical size (length) of a maximum transform block, the coding unit may be horizontally divided into halves.

[00205] The maximum and/or minimum coding unit size information and the maximum and/or minimum transform block size information may be signaled or determined at a higher coding unit level. The higher level may be, for example, a sequence level, a picture level, a slice level, a tile group level, a tile level, or the like. For example, the minimum coding unit size may be defined as 4x4. For example, the maximum transform block size may be defined as 64×64. For example, the minimum transform block size may be defined as 4x4.

[00206] Information about the minimum size (minimum quadtree size) of a coding unit corresponding to a leaf node of the quadtree, and/or information about the maximum depth (maximum tree depth for a tree of multiple types) from a root node to a leaf node of a tree of multiple types may be signaled or determined at higher level coding unit. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, a tile level, or the like. Minimum quadtree size information and/or maximum tree depth information of multiple types may be signaled or determined for each of the intra-picture slice and inter-picture slice.

[00207] The difference information between the CTU size and the maximum transform unit size may be signaled or determined at a higher coding unit level. For example, the higher level may be a sequence level, a picture level, a slice level, a tile group level, a tile level, or the like. Information about the maximum size of encoding units corresponding to the corresponding nodes of the binary tree (hereinafter referred to as the maximum size of the binary tree) may be determined based on the unit size of the encoding tree and the delta information. The maximum size of coding units corresponding to the corresponding nodes of the ternary tree (hereinafter referred to as the maximum size of the ternary tree) may vary depending on the type of slice. For example, for an intra-picture slice, the maximum size of the ternary tree can be 32x32. For example, for an inter-picture slice, the maximum size of the ternary tree can be 128x128. For example, the minimum size of encoding units corresponding to the corresponding binary tree nodes (hereinafter referred to as the minimum binary tree size) and/or the minimum size of encoding units corresponding to the corresponding ternary tree nodes (hereinafter referred to as the minimum ternary tree size) may be set as minimum coding block size.

[00208] As another example, the maximum binary tree size and/or the maximum ternary tree size may be signaled or determined at the slice level. Alternatively, the minimum binary tree size and/or the minimum ternary tree size may be signaled or determined at the slice level.

[00209] Depending on the size and depth information of the various blocks described above, quad-partitioning information, multiple-type tree partitioning indication information, partitioning tree information, and/or partitioning direction information may or may not be included in the bitstream.

[00210] For example, when the size of a coding unit is not larger than the minimum quadtree size, the coding unit does not contain quad-partitioning information. Thus, the quad-partition information can be inferred from the second value.

[00211] For example, when the dimensions (horizontal and vertical dimensions) of a coding unit corresponding to a tree node of a plurality of types are larger than the maximum dimensions (horizontal and vertical dimensions) of a binary tree and/or the maximum dimensions (horizontal and vertical dimensions) of a ternary tree, the unit encoding cannot be binary separated or ternary separated. Accordingly, tree splitting indication information of a plurality of types may not be signaled, but may be derived from the second value.

[00212] Alternatively, when the dimensions (horizontal and vertical dimensions) of the encoding unit corresponding to a tree node of the plurality of types are the same as the maximum dimensions (horizontal and vertical dimensions) of the binary tree and/or twice larger than the maximum dimensions (horizontal and vertical dimensions) of a ternary tree, the coding unit cannot be further binary-divided or ternary-divided. Accordingly, tree splitting indication information of a plurality of types may not be signaled, but may be derived from the second value. This is because when a coding unit is split according to a binary tree splitting structure and/or a ternary tree splitting structure, a coding unit smaller than the minimum binary tree size and/or the minimum ternary tree size is generated.

[00213] Alternatively, the binary tree split or ternary tree split may be limited based on the size of the virtual pipeline data unit (hereinafter, pipeline buffer size). For example, when an encoding unit is divided into sub-encoding units that do not correspond to a pipeline buffer size by a binary tree split or a ternary tree split, the corresponding binary tree split or ternary tree split may be constrained. The pipeline buffer size can be the size of the maximum translation block (for example, 64x64). For example, when the pipeline buffer size is 64x64, the division below may be limited.

[00214] - N×M (N and/or M is 128) Ternary tree partition for coding units

[00215] - 128×N (N<=64) Horizontal binary tree split for coding units

[00216] - N×128 (N<=64) Vertical binary tree partition for coding units

[00217] Alternatively, when the depth of a coding unit corresponding to a multi-type tree node is equal to the maximum depth of the multi-type tree, the coding unit cannot be further binary and/or ternary split. Accordingly, tree splitting indication information of a plurality of types may not be signaled, but may be derived from the second value.

[00218] Alternatively, only when at least one of vertical binary tree splitting, horizontal binary tree splitting, vertical ternary tree splitting, and horizontal ternary tree splitting is possible for a coding unit corresponding to a tree node of a plurality of types, tree splitting indicating information of a plurality of types may be signaled. Otherwise, the coding unit cannot be binary and/or ternary split. Accordingly, tree splitting indication information of a plurality of types may not be signaled, but may be derived from the second value.

[00219] Alternatively, only when, for a coding unit corresponding to a tree node of multiple types, both a binary tree split in the vertical direction and a binary tree split in the horizontal direction, or both a ternary tree split in the vertical direction and a ternary tree split in the horizontal direction, are possible horizontal direction, splitting direction information may be signaled. Otherwise, the split direction information may not be signaled, but may be inferred from a value indicating possible split directions.

[00220] Alternatively, only when, for a coding tree corresponding to a tree node of multiple types, both a binary tree split in the vertical direction and a ternary tree split in the vertical direction are possible, or both a binary tree split in the horizontal direction and a ternary tree split in the horizontal direction, partition tree information may be signaled. Otherwise, the split tree information may not be signaled, but may be inferred from a value indicating a possible split of the tree structure.

[00221] FIG. 4 is a view showing the intra-prediction process.

[00222] The arrows are from the center outward in FIG. 4 may represent prediction directions of intra-prediction modes.

[00223] Intra encoding and/or decoding may be performed using a reference sample of an adjacent block of the current block. The neighbor block may be a reconstructed neighbor block. For example, intra-encoding and/or decoding may be performed using an encoding parameter or a reference sample value included in the reconstructed neighboring block.

[00224] A prediction block may mean a block generated by performing intra-prediction. The prediction unit may correspond to at least one of a CU, a PU, and a TU. The prediction block unit may have the size of one of CU, PU and TU. The prediction block may be a square block having a size of 2x2, 4x4, 16x16, 32x32 or 64x64, etc., or may be a rectangular block having a size of 2x8, 4x8, 2x16, 4x16 and 8x16, etc.

[00225] Intra prediction may be performed according to the intra prediction mode for the current block. The number of intra-prediction modes that the current block can have may be a fixed value and may be a value defined differently according to an attribute of the prediction block. For example, the prediction block attribute may include the size of the prediction block and the shape of the prediction block, etc.

[00226] The number of intra-prediction modes may be fixed to N regardless of the block size. Or the number of intra-prediction modes can be 3, 5, 9, 17, 34, 35, 36, 65 or 67, etc. Alternatively, the number of intra-prediction modes may vary according to the block size or color component type or both. For example, the number of intra-prediction modes may vary according to whether the color component is a luma signal or a chrominance signal. For example, when the block size becomes large, the number of intra-prediction modes may increase. Alternatively, the number of intra-prediction modes of the luma component block may be greater than the number of intra-prediction modes of the chrominance component block.

[00227] The intra-prediction mode may be a non-angular mode or an angular mode. The non-angular mode may be a DC mode or a planar mode, and the angular mode may be a prediction mode having a specific direction or angle. The intra-prediction mode may be expressed by at least one of a mode number, a mode value, a mode number symbol, a mode angle, and a mode direction. The number of intra-prediction modes can be M, which is greater than 1, including non-angular and angular mode. For intra-prediction of the current block, the step of determining whether samples included in the reconstructed neighboring block can be used as reference samples of the current block can be performed. When there is a sample that cannot be used as the reference sample of the current block, the value obtained by duplicating or interpolating at least one sample value among the samples included in the reconstructed adjacent block, or both, can be used to replace the unused sample value for the sample. thus, the replaced sample value is used as the reference sample of the current block.

[00228] FIG. 7 is a diagram illustrating reference samples that can be used for intra-prediction.

[00229] As shown in FIG. 7, at least one of reference sample line 0 and reference sample line 3 may be used for intra-prediction of the current block. In fig. 7, the samples of segment A and segment F can be padded with samples closest to segment B and segment E, respectively, instead of extracting from the reconstructed neighboring block. Index information indicating the reference sample line that should be used for intra-prediction of the current block may be signaled. When the upper boundary of the current block is the CTU boundary, only line 0 of the reference sample can be accessed. Therefore, in this case, the index information may not be signaled. When a reference sample line other than reference sample line 0 is used, filtering for the prediction block to be described later may not be performed.

[00230] In intra-prediction, a filter may be applied to at least one of a reference sample and a prediction sample based on the intra-prediction mode and the current block size.

[00231] In the case of the planar mode, when generating a prediction block of the current block, according to the position of the prediction target sample in the prediction block, the sample value of the prediction target sample can be generated using the weighted sum of the top and left side reference samples for the current sample, and the reference samples the top right side and bottom left side of the current block. In addition, in the case of DC mode, when generating a prediction block of the current block, the average value of the reference samples of the top and left side of the current block may be used. Moreover, in the case of corner mode, a prediction block may be generated using a reference sample of the top side, left side, right top side, and/or left bottom side of the current block. To generate a prediction sample value, unit interpolation of the real number may be performed.

[00232] In the case of intra-prediction between color components, a prediction block for the current block of the second color component can be generated based on the corresponding reconstructed block of the first color component. For example, the first color component may be a luma component, and the second color component may be a chrominance component. For intra-prediction between color components, parameters of a linear model between the first color component and the second color component can be obtained based on the template. The template may include top and/or left neighbor samples of the current block and top and/or left neighbor samples of the reconstructed block of the first color component corresponding thereto. For example, the linear model parameters may be derived using the sample value of the first color component having the maximum value among the samples in the template, and the sample value of the second color component corresponding to it, and the sample value of the first color component having the minimum value among the samples in the template, and the sample value of the second color component corresponding to it. Once the parameters of the linear model are inferred, the corresponding reconstructed block can be applied to the linear model to generate a prediction block for the current block. According to the video format, subsampling may be performed on adjacent samples of the reconstructed block of the first color component and the corresponding reconstructed block. For example, when one sample of the second color component corresponds to four samples of the first color component, the four samples of the first color component may be subsampled to compute one corresponding sample. In this case, inference of linear model parameters and intra-prediction between color components can be performed based on the corresponding subsampled samples. Whether or not to perform intra-prediction between color components and/or pattern range may be signaled as an intra-prediction mode.

[00233] The current block can be divided into two or four sub-blocks in the horizontal or vertical direction. Broken subblocks can be sequentially restored. That is, intra-prediction may be performed on the sub-block to generate a prediction sub-block. Additionally, dequantization and/or deconversion may be performed on the subblocks to generate residual subblocks. The reconstructed subblock can be generated by adding a prediction subblock to the residual subblock. The reconstructed sub-block can be used as a reference sample for intra-prediction of sub-sub-blocks. A subblock may be a block containing a predetermined number (eg, 16) or more samples. Accordingly, for example, when the current block is an 8×4 or 4×8 block, the current block may be divided into two sub-blocks. In addition, when the current block is a 4x4 block, the current block cannot be divided into subblocks. When the current block has different sizes, the current block can be split into four subblocks. Information about whether to perform intra-prediction based on sub-blocks and/or the direction of partitioning (horizontal or vertical) may be signaled. Intra-prediction based on sub-blocks can be limited to execution only when reference sample line 0 is used. When intra-prediction based on a sub-block is performed, filtering for the prediction block, which will be described later, may not be performed.

[00234] The final prediction block can be generated by performing filtering on a prediction block that is intra-predicted. Filtering can be performed by applying predefined weights to the filter target sample, the left reference sample, the top reference sample, and/or the top left reference sample. The weight and/or reference sample (range, position, etc.) used for filtering may be determined based on at least one of block size, intra-prediction mode, and position of the target filter sample in the prediction block. Filtering can only be performed in the case of a predefined intra-prediction mode (eg, DC, planar, vertical, horizontal, diagonal and/or adjacent diagonal modes). An adjacent diagonal mode may be a mode in which k is added to or subtracted from the diagonal mode. For example, k can be a positive integer 8 or less.

[00235] The intra-prediction mode of the current block can be entropy encoded/decoded by predicting the intra-prediction mode of a block adjacent to the current block. When the intra-prediction modes of the current block and the neighboring block are identical, information that the intra-prediction modes of the current block and the neighboring block are identical may be signaled using predefined flag information. In addition, intra-prediction mode indicator information that is identical to the intra-prediction mode of the current block among the intra-prediction modes of a plurality of neighboring blocks may be signaled. When the intra-prediction modes of the current block and the neighboring block are different, the intra-prediction mode information of the current block can be entropy encoded/decoded by performing entropy encoding/decoding based on the intra-prediction mode of the neighboring block.

[00236] FIG. 5 is a diagram illustrating an embodiment of the inter-prediction process.

[00237] In FIG. 5 a rectangle can represent a picture. In fig. 5, the arrow represents the direction of the prediction. Pictures can be classified into intrinsic pictures (I-pictures), predictive pictures (P-pictures) and bi-predictive pictures (B-pictures) according to their encoding type.

[00238] The I-picture can be encoded by intra-prediction without the need for inter-picture (inter-frame) prediction. The P-picture may be encoded by inter-picture prediction using a reference picture that is present in one direction (ie, forward direction or backward direction) relative to the current block. The B-picture may be encoded by inter-picture prediction using reference pictures that are preset in two directions (ie, forward direction and backward direction) relative to the current block. When inter-picture prediction is used, the encoder may perform inter-prediction or motion compensation, and the decoder may perform corresponding motion compensation.

[00239] An embodiment of inter-picture prediction will be described in detail below.

[00240] Inter-prediction or motion compensation may be performed using the reference picture and motion information.

[00241] Motion information of the current block can be output during inter-prediction by each of the encoding device 100 and the decoding device 200. The motion information of the current block can be output by using the motion information of a reconstructed adjacent block, the motion information of a combined block (also called a col block or a collocated block), and/or a block adjacent to a collocated block. A aligned block may mean a block that is spatially located at the same position as the current block in a previously reconstructed fused picture (also referred to as a col picture or fused picture). The combined picture may be one picture from one or more reference pictures included in the list of reference pictures.

[00242] The method for outputting motion information may differ depending on the prediction mode of the current block. For example, the prediction mode used for inter-prediction includes AMVP mode, fusion mode, skip mode, motion vector difference fusion mode, sub-block fusion mode, triangular partition mode, inter-intra combined prediction mode, affine mode, etc. P. Here, the fusion mode may be referred to as the motion fusion mode.

[00243] For example, when AMVP is used as a prediction mode, at least one of motion vectors of reconstructed adjacent blocks, motion vectors of aligned blocks, motion vectors of blocks adjacent to aligned blocks, and a (0, 0) motion vector may be determined in as motion vector candidates for the current block, and a list of motion vector candidates is generated using the motion vector candidates. The motion vector candidate of the current block can be inferred using the generated list of motion vector candidates. The motion information of the current block can be determined based on the derived motion vector candidate. Motion vectors of aligned blocks or motion vectors of blocks adjacent to aligned blocks may be referred to as temporal motion vector candidates, and motion vectors of reconstructed adjacent blocks may be referred to as spatial motion vector candidates.

[00244] The encoding apparatus 100 may calculate a motion vector difference (MVD) between a motion vector of the current block and a motion vector candidate, and may perform entropy encoding of the motion vector difference (MVD). In addition, the encoding device 100 can perform entropy encoding of the motion vector candidate index and generate a bitstream. The motion vector candidate index may indicate an optimal motion vector candidate among the motion vector candidates included in the motion vector candidate list. The decoding apparatus may perform entropy decoding on a motion vector candidate index included in the bitstream, and may select a motion vector candidate of the decoding target block from among the motion vector candidates included in the motion vector candidate list using the entropy decoded motion vector candidate index. In addition, the decoding apparatus 200 may add the entropy decoded MVD and the motion vector candidate extracted by entropy decoding, thereby outputting the motion vector of the decoding target block.

[00245] Meanwhile, the encoding device 100 can perform entropy encoding on the resolution information of the calculated MVD. The decoding apparatus 200 may adjust the resolution of the entropy-decoded MVD using the MVD resolution information.

[00246] Meanwhile, the encoding device 100 calculates the motion vector difference (MVD) between the motion vector and the motion vector candidate in the current block based on the affine model and performs entropy encoding on the MVD. The decoding apparatus 200 outputs a motion vector for each sub-block by deriving the affine control motion vector of the decoding target block through the sum of the entropy decoded MVD and the affine control motion vector candidate.

[00247] The bitstream may include a reference picture index indicating the reference picture. The reference picture index may be entropy encoded by encoder 100 and then signaled as a bitstream to decoder 200. The decoding apparatus 200 may generate a prediction block from a target decoding block based on the derived motion vector and reference picture index information.

[00248] Another example of a method for outputting current motion information may be a merge mode. Merge mode may refer to a method of merging the movement of multiple blocks. The merging mode may mean a mode of deriving motion information of the current block from motion information of neighboring blocks. When the merging mode is applied, a merging candidate list may be generated using motion information of reconstructed neighboring blocks and/or motion information of merged blocks. The motion information may include at least one of a motion vector, a reference picture index, and an inter-prediction indicator. The prediction indicator may indicate unidirectional prediction (L0 prediction or L1 prediction) or bidirectional prediction (L0 prediction and L1 prediction).

[00249] The merge candidate list may be a list of stored motion information. The motion information included in the fusion candidate list may be at least one of the motion information (spatial fusion candidate) of a neighboring block adjacent to the current block, the motion information (temporal fusion candidate) of the merged block of the current block in the reference picture, new motion information, generated by a combination of the motion information output from the merge candidate list, the motion information (history-based merge candidate) of the block that is encoded/decoded before the current block, and the null merge candidate.

[00250] The encoding apparatus 100 may generate a bitstream by performing entropy encoding on at least one of a merge flag and a merge index, and may signal the bitstream to the decoding apparatus 200. The merge flag may be information indicating whether to perform a merge mode for each block, and the merge index may be information indicating which neighboring block among neighboring blocks of the current block is a merge target block. For example, the neighbor blocks of the current block may include a left neighbor block on the left side of the current block, a top neighbor block located above the current block, and a temporary neighbor block temporally adjacent to the current block.

[00251] Meanwhile, the encoding device 100 performs entropy encoding on the correction information for correcting the motion vector among the motion information of the fusion candidate and signals it to the decoding device 200. The decoding apparatus 200 may correct the motion vector of the fusion candidate selected by the fusion index based on the correction information. Here, the correction information may include at least one of information about whether to perform correction or not, correction direction information, and correction size information. As described above, a prediction mode that corrects the motion vector of a fusion candidate based on the signaled correction information may be referred to as a fusion mode having a motion vector difference.

[00252] The skip mode may be a mode in which the motion information of a neighboring block is applied to the current block as is. When the skip mode is applied, the encoding device 100 can perform entropy encoding of information that the motion information of a particular block should be used as the motion information of the current block to generate a bit stream, and can signal the bit stream to the decoding device 200. The encoding device 100 may not signal a syntax element regarding at least any one of the motion vector difference information, the encoding block flag, and the level of transform coefficients to the decoding device 200.

[00253] A subblock merging mode may mean a mode that outputs motion information in units of coding unit (CU) subblocks. When the sub-block merging mode is applied, a sub-block merging candidate list can be generated using the motion information (sub-block-based temporary merging candidate) of the sub-block aligned with the current sub-block in the reference image and/or the affine control point motion vector merging candidate.

[00254] The triangular partitioning mode may mean a mode that outputs motion information by dividing the current block into diagonal directions, outputs each prediction sample using each output motion information, and extracts the prediction sample of the current block by weighting each of the output prediction samples.

[00255] A combined inter-intra prediction mode may mean a mode that outputs a prediction sample of the current block by weighting the prediction sample generated by inter-prediction and the prediction sample generated by intra-prediction.

[00256] The decoding device 200 can independently correct the output motion information. The decoding apparatus 200 can search a specified area based on the reference block indicated by the output motion information, and output motion information having a minimum SAD as corrected motion information.

[00257] The decoding apparatus 200 may compensate the prediction sample output by inter-prediction using optical flow.

[00258] FIG. 6 is a diagram illustrating the conversion and quantization process.

[00259] As shown in FIG. 6, a conversion and/or quantization process is performed on the residual signal to generate a quantized level signal. The residual signal represents the difference between the original block and the prediction block (ie, intra-prediction block or inter-prediction block). A prediction block is a block generated by intra-prediction or inter-prediction. The transformation may be a primary transformation, a secondary transformation, or both. Primary transformation of the residual signal results in transformation coefficients, and secondary transformation of the transformation coefficients results in secondary transformation coefficients.

[00260] At least one scheme selected from various conversion schemes that are predefined is used to perform the primary conversion. For example, examples of predefined transform schemes include discrete cosine transform (DCT), discrete sine transform (DST), and Karhunen-Loeve transform (KLT). Transformation coefficients generated by the primary transformation may be subject to a secondary transformation. The transform schemes used for the primary transform and/or secondary transform may be determined according to the encoding parameters of the current block and/or neighboring blocks of the current block. Alternatively, conversion information indicating the conversion scheme may be signaled. The DCT-based conversion may include, for example, DCT-2, DCT-8, and the like. The DST-based conversion may include, for example, DST-7.

[00261]

[00262] The quantized level signal (quantization coefficients) can be generated by performing quantization on the residual signal or the result of performing the primary transform and/or secondary transform. The quantized level signal may be scanned according to at least one of a diagonal up-right scan, a vertical scan, and a horizontal scan, depending on the block intra-prediction mode or the block size/shape. For example, when coefficients are scanned in a diagonal up-to-right scan, the coefficients in block form are converted to one-dimensional vector form. In addition to the up-right diagonal scan, horizontal scan horizontal scan of the 2D shape of the coefficient block or vertical scan of the vertical scan of the 2D shape of the coefficient block can be used depending on the intra-prediction mode and/or the size of the transform block. The scanned quantized layer coefficients can be entropy encoded for insertion into the bit stream.

[00263] The decoder performs entropy decoding of the bit stream to obtain quantized level coefficients. The quantized level coefficients can be arranged in a two-dimensional block form by backscanning. The reverse scan may use at least one of a diagonal up-right scan, a vertical scan, and a horizontal scan.

[00264] The quantized level coefficients can then be dequantized, then subjected to a secondary inverse transform if necessary, and finally a primary-inverse transform if necessary to generate a reconstructed residual signal.

[00265] Dynamic range inverse mapping may be performed on the luminance component reconstructed by intra-prediction or inter-prediction before in-loop filtering. The dynamic range can be divided into 16 equal parts, and the display function for each part can be signaled. The mapping function can be signaled at the slice level or the tile group level. The inverse mapping function for performing the inverse mapping can be derived based on the mapping function. Contour filtering, reference picture storage and motion compensation are performed in the inverse mapping region, and the prediction block generated by inter-prediction is converted into the mapping region by mapping using a mapping function, and then used to generate the reconstructed block. However, since intra-prediction is performed on the mapped area, the prediction block generated by intra-prediction can be used to generate a reconstructed block without mapping/demapping.

[00266] When the current block is a residual block of a chroma component, the residual block can be converted into a reverse display area by performing scaling of the chrominance component of the displayed area. Scaling availability can be signaled at the slice level or the tile group level. Scaling can only be applied when mapping is available for the luma component, and the luma component division and the chrominance component division follow the same tree structure. The scaling may be performed based on the average of the samples of the luminance prediction block corresponding to the color difference block. In this case, when the current block uses inter-prediction, the luminance prediction block may mean the mapped luminance prediction block. The value required for scaling can be obtained by accessing the lookup table using the index of the slice to which the average value of the brightness prediction block samples belongs. Finally, by scaling the residual block using the output value, the residual block can be switched to the reverse mapping region. Then, chrominance component block recovery, intra-prediction, inter-prediction, contour filtering, and reference picture storage can be performed in the demapping region.

[00267] Information indicating whether luma component and chrominance component mapping/demapping is available may be signaled through a set of sequence parameters.

[00268] A current block prediction block may be generated based on a block vector indicating an offset between the current block and a reference block in the current image. Thus, the prediction mode for generating a prediction block with reference to the current picture is called intra-block copy (IBC) mode. The IBC mode can be applied to M×N coding units (M<=64, N<=64). The IBC mode may include a skip mode, a merge mode, an AMVP mode, and the like. In the case of skip mode or merge mode, a list of merge candidates is generated and the merge index is signaled so that one merge candidate can be specified. The block vector of the specified merge candidate can be used as the block vector of the current block. The list of merge candidates may include at least one of a spatial candidate, a history-based candidate, a candidate based on the average of two candidates, and a null merge candidate. In the case of AMVP mode, a difference block vector may be signaled. In addition, the prediction block vector can be derived from the left neighbor block and the top neighbor block of the current block. The index for the adjacent block to be used may be signaled. A prediction block in IBC mode is included in the current CTU or left CTU and is limited to a block in the already reconstructed region. For example, the value of the block vector may be constrained such that the prediction block of the current block is located in a region of three 64x64 blocks preceding the 64x64 block to which the current block belongs, in encoding/decoding order. By limiting the block vector value in this way, memory consumption and device complexity in accordance with the IBC mode implementation can be reduced.

[00269]

[00270] Next, an image encoding/decoding method according to the present invention will be described based on the above description.

[00271] The present invention will describe a method for generating a candidate list for performing motion prediction. In a further embodiment, the candidate list may be an adaptive motion vector appearance list (AMVOL). As another example, in the following embodiment, the candidate list may be a history-based motion vector predictor (HMVP) list. An encoder or decoder can perform motion prediction more accurately and efficiently by generating and using a candidate list in accordance with the present invention, thereby improving coding efficiency.

[00272] In accordance with the present invention, when performing motion prediction, an encoder or decoder may perform motion prediction using motion information obtained in a predetermined section or region.

[00273] The motion information may be stored and used in a block having a higher level than the block in which motion information prediction is performed. An example of a unit in which motion prediction is performed includes a CU, a PU, a block, a macroblock, and the like. The unit having a higher level than the unit in which the motion information prediction is performed may include all units larger than the unit in which the motion information prediction is performed. For example, a unit having a higher level here may include at least one of a block, CTU, brick, tile, slice, sub-picture, picture, frame, and group of pictures (GOP) consisting of one or more C.U.

[00274] In the following embodiments, a higher unit larger than the unit in which motion information prediction is performed may be represented by a higher prediction unit (UPU). That is, in the following embodiments, a UPU may mean at least one of a block, a CTU, a brick, a tile, a slice, a sub-picture, a picture, a frame, and a group of pictures (GOP) consisting of one or more CUs.

[00275] FIG. 8 is a diagram illustrating a higher upper prediction unit (UPU) used in some embodiments of the present invention.

[00276] As used herein, a UPU may refer to a set of units in which motion prediction is performed and a structure having one or more candidate lists. Further, a block belonging to a particular UPU can be expressed as a block having a lower level than the corresponding UPU, or a block included in the UPU. A candidate list for a particular UPU can be constructed using motion information generated from blocks having a lower level than the UPU. The candidate list may be configured to store all or part of the motion information generated from blocks having a lower level than the UPU. The sizes or units of different UPUs may be the same or may be installed differently.

[00277] With reference to FIG. 8 describes a UPU that may be used in some embodiments of the present invention. Picture in Fig. 8 includes a plurality of encoding blocks.

[00278] FIG. 8(a) illustrates the case where many pictures belong to one UPU. That is, fig. 8(a) illustrates the case where the UPU is a GOP unit. In this case, the UPU may have a candidate list generated using motion information generated from pictures included in the UPU.

[00279] FIG. 8(b) illustrates the case where one picture belongs to one UPU. In this case, the UPU may have a candidate list generated using the motion information generated in each image.

[00280] FIG. 8(c)-8(e) illustrate cases in which one picture includes multiple UPUs. That is, fig. 8(c)-8(e) illustrate cases in which the UPU is a unit having a lower level than the picture. For example, the UPU can be set by performing horizontal quarter-splitting of a picture (FIG. 8(c)) or performing vertical quarter-splitting of a picture. As another example, the UPU can be established by performing n-splitting of the picture.

[00281] Meanwhile, FIG. 8(d) may be an example in which the UPU consists of CTUs forming a single row. That is, the UPU can be installed in the CTU line unit. Here, the UPU may have a candidate list generated using the traffic information generated in each CTU line.

[00282] Although not shown in FIG. 8, UPU can be installed in a CTU and can be installed in a slice, brick and multiple CUs.

[00283] Further, the motion information included in the candidate list may include at least one of information necessary for motion prediction, such as a motion vector, a reference picture number (index, identifier, etc.), a reference list pictures, motion prediction direction, etc.

[00284] The unit in which motion information generated in blocks included in the UPU is stored may be called a candidate list. The motion information stored in the candidate list can be used when the encoder or decoder is operating in inter-prediction mode or IBC mode. The candidate list may include additional information that can be used to use the candidate list in addition to the traffic information. Additional information that may be included in the candidate list may include, for example, information about the frequency of occurrence of motion information, information about a position where motion information appears, and the like.

[00285] FIG. 9 is a diagram illustrating the structure of an encoder in accordance with an embodiment of the present invention.

[00286] Referring to FIG. 9, motion information generated as a result of inter-prediction can be added to a candidate list in accordance with the present invention. As another example, although not shown in FIG. 9, motion information generated as a result of IBC prediction can be added to a candidate list in accordance with the present invention. The candidate list can then be used for inter-prediction or IBC, thereby improving coding efficiency.

[00287] FIG. 10 is a diagram illustrating a structure of a decoder according to an embodiment of the present invention.

[00288] Referring to FIG. 10, motion information recovered by inter-prediction can be added to a candidate list in accordance with the present invention. As another example, although not shown in FIG. 10, motion information recovered by IBC prediction can be added to a candidate list in accordance with the present invention. The candidate list can then be used again in inter-prediction or IBC to recover motion information for the current block.

[00289] FIG. 11 is a flowchart illustrating an encoding/decoding method in accordance with an embodiment of the present invention.

[00290] FIG. 11(a) shows a coding method of an encoder to which the present invention is applied. General encoder motion information prediction may refer to a motion information prediction method performed by an encoder according to the prior art. Predicting motion information in a candidate list may mean predicting motion information using a candidate list in accordance with the present invention. The motion prediction step means the motion prediction step of the current block, and the result of motion information prediction can be used in this process. Selecting the optimal prediction may mean selecting the prediction most optimal for encoding from the motion information prediction results derived from the general motion information prediction and the motion information prediction results derived from the motion information prediction using the candidate list. Encoding the motion information may mean a step of an encoding method according to a selection result of the optimal prediction, additional information related thereto, and information related to the motion information. Adding motion information from the candidate list may include adding previously generated motion information to the candidate list. The motion information added by adding motion information from the candidate list can be used later in the encoder's inter-prediction.

[00291] FIG. 11(b) shows a decoding method of a decoder to which the present invention is applied. Decoding the motion information may be a step of decoding the signal encoded by the encoder. The decoded signal may include a decoding method for decoding the current block, additional information about the decoding method, and information related to motion information. General motion information prediction may refer to a motion information prediction method performed in a general decoder. Predicting motion information of a candidate list may mean predicting motion information using the candidate list in accordance with the present invention. There are many results of general motion information prediction and candidate list motion information prediction. The decoder can determine whether the optimal prediction is motion information prediction using the candidate list, thereby selecting a motion information prediction method to be used for decoding among the candidate list motion information prediction method and the general motion information prediction method. A method for predicting motion information can be determined by information signaled by the encoder. Motion compensation may mean a step of decoding motion information of an actual block, and may mean a step of decoding motion information of a current block using the decoded signal and a prediction result of the motion information. Adding the candidate list motion information may be a step of adding the finally decoded motion information to the candidate list. The motion information added by adding candidate list motion information can be used later in the inter-prediction of the decoder.

[00292] FIG. 12 is a diagram illustrating structures of an encoder and decoder in accordance with another embodiment of the present invention.

[00293] In FIG. 12(a) shows the configuration of an encoder to which the present invention is applied. The general motion information prediction unit may mean a motion information prediction unit performed by the encoder. The candidate list motion information prediction unit may relate to a configuration for performing motion information prediction using motion information stored in the candidate list according to the present invention. There are a plurality of motion information prediction results of the general motion information prediction module and the candidate list motion information prediction module. The motion prediction unit refers to a configuration for predicting the motion of a current block, and the result of motion information prediction can be used in this process. The optimal prediction selecting unit may be configured to select a prediction optimal for encoding the current block among the motion information prediction results derived from the general motion information prediction and the motion information prediction results derived from the motion information prediction of the candidate list. The motion information encoding unit may be configured to encode an encoding method according to a selection result of the optimal prediction selection unit, additional information related thereto, and information related to the motion information. The candidate list adding module may be configured to add previously generated motion information to the candidate list. The motion information added by the candidate list adding module can be used later in inter-prediction or by the IBC encoder.

[00294] FIG. 11(b) is a diagram showing a decoder inter-prediction apparatus to which the present invention is applied. The motion information decoding module may be configured to decode a signal encoded by the encoder. The decoded signal may include a decoding method for decoding the current block, additional information about the decoding method, and information related to motion information. The general motion information prediction unit may refer to a configuration that executes a motion information prediction method performed by a conventional decoder. The candidate list motion information prediction unit may relate to a configuration for performing motion information prediction using motion information stored in the candidate list. There are many results of general motion information prediction and candidate list motion information prediction. The motion compensator can decode the motion information of the current block using the signal decoded by the motion information decoding unit and the motion information prediction result. The candidate list adding module may be configured to add the finally decoded motion information to the candidate list. The motion information added by the candidate list adding module can be used later in the inter-prediction or IBC of the decoder.

[00295] When new motion information or a candidate is stored in the candidate list, the same motion information or candidate may not be stored in the candidate list. The encoder or decoder may determine whether motion information overlapping with the new motion information is included in the current candidate list. An encoder or decoder can perform different operations according to whether they overlap.

[00296] Here, the criterion for determining the same motion information may be related to whether all the information included in the motion information is the same and whether some parts thereof are the same. For example, given that motion information consists of a motion vector and a reference picture number, when only the motion vector is compared and the motion vectors are the same, the encoder or decoder may determine that the motion information to be added is the same as traffic information included in the candidate list. In addition, the encoder or decoder can determine that the motion information to be added is the same as the motion information included in the candidate list only when both the motion vector and the reference picture number are the same.

[00297] The motion information from the candidate list in accordance with the present invention can determine the prediction potential of the motion information, and store and update information representing the same. Here, the candidates from the candidate list can be ordered according to the motion information prediction potential. Next, a method for determining the prediction potential of a candidate list will be described.

[00298] The size of the candidate list in accordance with the present invention may be limited. Here, the size of the candidate list may mean the maximum traffic information or the number of candidates that can be stored in the candidate list. When the number of pieces of motion information stored in the candidate list reaches the maximum value, the encoder or decoder may add (store) or skip storing new motion information in the candidate list according to the following example.

[00299] The size of the candidate list may be limited to a preset value in the encoder or decoder. For example, the preset value may be a value from 1 to 6. As another example, the size of the candidate list may be limited by a value signaled from the encoder to the decoder. Information about values for limiting the size of the candidate list may be signaled at a higher bitstream level, such as a sequence parameter set, a picture parameter set, a slice header, a tile header, and a tile group header.

[00300] For example, when the number of pieces of motion information stored in the candidate list reaches a maximum value, the newly outputted motion information may be omitted without being stored in the candidate list.

[00301] As another example, when the number of pieces of motion information stored in the candidate list reaches a maximum value, the prediction potential of the newly outputted motion information is compared with the prediction potential of motion information stored in the existing candidate list, so that motion information with high prediction potential is added to the candidate list, and the motion information with the lowest prediction potential can be removed from the candidate list. Here, when the prediction potential of the newly outputted motion information is lower than the prediction potential of the motion information stored in the existing candidate list, the newly outputted motion information may be omitted without being stored in the candidate list.

[00302] As another example, when the number of pieces of motion information stored in the candidate list reaches the maximum value, the motion information stored first among the motion information stored in the candidate list is deleted, and the newly outputted motion information can be added to the candidate list . That is, the newly generated motion information can be added to the candidate list using a first-in, first-out (FIFO) rule.

[00303] That is, when the number of pieces of motion information stored in the candidate list reaches the maximum value, the motion information indicated by index 0 of the candidate list can be deleted, and the newly outputted motion information can be added to the candidate list. Here, the index of motion information included in the candidate list may be reduced by one. Newly added motion information may be added to the candidate list and is indicated by the largest index among the motion information included in the candidate list.

[00304] The process of adding motion information to the candidate list may be omitted.

[00305] For example, when the same motion information is already stored in the candidate list, the newly outputted motion information may not be added to the candidate list. As another example, when similar motion information is already stored in the candidate list, the newly outputted motion information may not be added to the candidate list. As another example, when the number of pieces of motion information stored in the candidate list has already reached the maximum, the newly outputted motion information may not be added to the candidate list. As another example, when the prediction potential of motion information to be stored is low, newly derived motion information may not be added to the candidate list.

[00306] As another example, when the motion information to be newly added is already stored in the candidate list, the previously stored overlapping motion information can be removed from the candidate list, and the newly outputted motion information can be added to the candidate list. Here, the motion information indexes of the candidate list having an index value greater than the deleted motion information may be reduced by one. The newly added motion information may be added to the candidate list and is indicated by the largest index among the motion information included in the candidate list.

[00307] Here, that the motion information is similar may mean a case where the position difference of the reference area indicated by the motion information is equal to or less than a predetermined threshold value. The motion vector can be expressed as a combination of (dx, dy) the x-axis displacement dx and the y-axis displacement dy. dx represents how far the reference region is separated from the current block's x-axis position, and dy represents how far the reference region is separated from the current block's y-axis position. For example, MV1 is expressed as (2, 3), MV2 is expressed as (3, 3), MV3 is expressed as (10, 5), and the predefined threshold value can be 3 on the x-axis and 3 on the y-axis. Here, since MV1 and MV2 have a difference of 1 in the x-axis and 0 in the y-axis, all of which are less than or equal to the threshold value, MV1 and MV2 can be determined to have motion information similar to each other. Meanwhile, since MV1 and MV3 have a difference of 8 in the x-axis and 2 in the y-axis, which exceeds the threshold in the x-axis, it is determined that MV1 and MV3 are not similar motion information.

[00308] The process of comparing the difference between a threshold value and a motion information pair and the process of comparing the difference between a motion information pair and the difference between another motion information pair may be referred to as motion information similarity comparison. Here, similarity comparison may be performed on all information that is contained in the motion information, and similarity comparison may be performed on only some information from the motion information. For example, when the motion information consists of information that shows a motion vector and a reference picture, a similarity comparison can be performed only on the motion vector, and a similarity comparison can be performed by comparing all the information showing a motion vector and the reference picture.

[00309] The motion information stored in the candidate list can be used as a candidate for motion information prediction. That is, the motion information can be used as a candidate for all motion information prediction methods that predict current motion information with reference to other motion information. For example, motion information stored in the candidate list can be used as one of the motion information candidates when the encoder or decoder operates in the merging mode, and the encoder or decoder operates in the AMVP mode or the IBC mode.

[00310] When using the motion information stored in the candidate list as a motion information prediction candidate, the motion information used as the candidate can be determined as a candidate of the candidate list according to the present invention.

[00311] The candidate used for conventional motion information prediction and the candidate in the candidate list can be used simultaneously, and motion information prediction can be performed using only the candidates in the candidate list.

[00312] For example, when a candidate list candidate is used together with a regular motion information prediction candidate, candidates in the candidate list may be generated excluding motion information from the candidate list having a higher priority than the candidate from the candidate list or having the same motion information , the same as the previously generated candidate. Here, priority may mean the priority of motion information candidates referred to in inter-prediction or IBC. Since the priority of the motion information candidate is higher, the encoder can perform signaling by assigning fewer bits to the candidate.

[00313] In addition, when configuring the AMVP or merge candidate list, when the size of the AMVP or merge candidate list is limited, the AMVP or merge candidate list can be configured according to priority. For example, starting with a candidate with a higher priority, it can be added to the AMVP or merge candidate list. When candidate motion information is stored in the candidate list up to the maximum size of the AMVP or merge candidate list, the low priority candidate is not added to the candidate list and cannot be used to predict motion information.

[00314] In addition, when the AMVP or merge candidate list is configured using high priority traffic information, the encoder or decoder may signal information indicating a particular candidate.

[00315] FIG. 13 is a diagram illustrating a list of candidates in accordance with an embodiment of the present invention.

[00316] FIG. 13 illustrates a case where an embodiment of the present invention is applied to a merge mode and an AMVP mode.

[00317] In the merge candidate list, the priority may be set in the order of spatial candidate, temporal candidate, mixed list candidate, and zero vector candidate. Spatial candidates can have priority in the order A1, B1, B0, A0 and B2. The temporary candidate may have priority in the order of H and C, and at least one of the two candidates may be added to the merge candidate list. A mixed list candidate for bidirectional prediction can be generated by combining a spatial candidate and a temporal candidate added to the candidate list. The null vector candidate may generate a candidate having a motion vector as the null vector indicating that there is no motion.

[00318] Accordingly, a merge candidate may have priority in the order A1, B1, B0, A0, B2, H or C, a mixed list candidate and a zero vector candidate, and candidates available in high priority order may be added to the merge candidate list . However, when candidates are added to the maximum size of the candidate list, no more candidates can be added to the candidate list.

[00319] Here, candidates from the candidate list in accordance with the present invention can be added to the merge candidate list. Candidates added to the merge candidate list can be located in any position, such as front, back, middle, etc. candidates not using the candidate list according to the present invention, and multiple candidates can be added through different priorities. Fig. 13 illustrates a situation where candidates from two candidate lists are added to the merge candidate list after the spatial candidate and the temporal candidate. Candidates in the candidate list may be added in various positions, for example, in front of spatial candidates, between spatial candidates, in front of temporary candidates, behind temporary candidates, and may also be added in two or more positions. Here, candidates from one or more candidate lists can be added to the merge candidate list.

[00320] The candidate list can be configured by applying various methods described herein using the motion information of all blocks encoded before the current block. In addition, by subsampling the motion information of blocks in the picture (for example, selecting the motion information of the blocks including the M-th motion information or the position of the M-th pixel), or by subsampling the position of the pixel in the picture at regular intervals and selecting the motion information of the block including corresponding pixel position, various methods described herein are applied using the selected motion information, thereby configuring the candidate list.

[00321] Here, the motion information of the first coded block in one picture is referred to as the first motion information, and the motion information of the M-th coded block is referred to as the M-th motion information. M can be a positive integer greater than zero.

[00322] Some or all of the traffic information included in the candidate list configured as described above may be selected as candidate list candidates according to various methods included herein and then added to the merge candidate list. Alternatively, the candidate list for motion information prediction may be configured using only the candidates of the candidate list in accordance with the present invention, and an index indicating the candidate needed to decode the corresponding block may be signaled.

[00323] To encode blocks included in the K-th CTU line to be currently encoded, movement information of blocks adjacent to the lower boundary of the K-1th CTU line can be used as candidates in merge mode and AMVP. To do this, the motion information and the position information of the corresponding blocks must be stored in a buffer (hereinafter called a row buffer) (K is any integer greater than 1).

[00324] For example, by using candidates from the candidate list in accordance with the present invention, rather than the corresponding motion information (information about a neighboring block located on a CTU line boundary), the encoder or decoder will not use a line buffer. In other words, when using the motion information included in the K-1st CTU line among the blocks included in the K-th CTU line as a merge candidate or AMVP, the corresponding motion information is not used, and the candidate list from the candidate list is used in accordance with the present invention.

[00325] For example, candidates from the candidate list may only store movement information of blocks adjacent to the lower boundary of the K-1st CTU row. In addition, only some of the movement information of blocks adjacent to the lower boundary of the K-1st row of the CTU may be stored. For example, only block motion information including the position of the Nth pixel to the left of the picture may be included in the candidate list. Here, N is an integer equal to 4 or more, and can be 4, 8, 16, 32, 64 or 128.

[00326] In AMVP mode, priority can be set in the order of spatial candidate, temporal candidate, and zero vector candidate. One candidate can be selected in region A and region B as a spatial candidate. Region A may mean candidates A0 and A1 and has priority in the order of A0, A1, scaled A0 and scaled A1, and at least one candidate can be selected among them. Here, the scaled candidate refers to a candidate generated by motion vector correction when the spatial motion vector reference picture is different from the block reference picture currently searched. Region B may indicate candidate B0, B1, or B2, and at least one candidate of B0, B1, B2, scaled B0, scaled B1, and scaled B2 may be selected. The temporary candidate may have priority order H and C, and at least one of the two may be added to the AMVP candidate list. The null vector candidate may generate a candidate having a motion vector as a null vector indicating that there is no motion.

[00327] Therefore, a candidate may have the priority of region candidate A, region candidate B, H or C, vector zero candidate, and the available candidates may be added to the AMVP candidate list in order of high priority. Here, once candidates are added up to the maximum AMVP candidate list size, no more candidates can be added to the candidate list.

[00328] Here, candidates of the candidate list in accordance with the present invention can be added to the AMVP candidate list. The candidates of the candidate list according to the present invention added to the AMVP candidate list may be located in any positions, such as before, after, in the middle, etc. a candidate that does not use the candidate list in accordance with the present invention, and a plurality of candidates may be added to the AMVP candidate list as priorities different from each other. Fig. 13 illustrates a situation in which candidates of two candidate lists are added to the AMVP candidate list after the spatial candidate and the temporal candidate. Candidates in the candidate list may be added in various positions, for example, before spatial candidates, between spatial candidates, before temporal candidates, after temporal candidates, etc. or may be added in two or more positions. In addition, candidates in one or more candidate lists may be added to the AMVP candidate list.

[00329] As another example, candidates of the candidate list in accordance with the present invention may be added to the candidate list for IBC prediction. The candidate list candidates added to the IBC candidate list in accordance with the present invention may be located in any position, such as before, after, in the middle, etc. candidates who do not use the candidate list in accordance with the present invention, and multiple candidates have different priorities and can be added to the IBC candidate list. In addition, candidates from one or more candidate lists may be added to the IBC candidate list.

[00330] FIG. 14 is a diagram illustrating a method for adding motion information to a candidate list in accordance with one embodiment of the present invention.

[00331] FIG. 14 shows a process in which motion information generated as a prediction is added to a candidate list. In fig. 14, (a), (b), (c), (d), (e), (f), (g), (h) and (i) mean the time order of the encoding/decoding process, encoding/decoding occurs in direction of the arrow between steps. Each rectangular area may indicate a coding unit or a prediction unit, and may indicate a unit in which prediction is performed so that motion information is generated. The gray encoding block may refer to the current block in which encoding/decoding is being performed at the current time in each drawing. In fig. 14, the entire area means a UPU, and the UPU may be composed of a set of coding units. The candidate list for each stage is shown at the top, and each column of the candidate list can mean stored motion information. Here, the number shown in the candidate list may indicate the index value of the candidate list. Colored information among the motion information stored in the candidate list may indicate newly added motion information at a corresponding stage. Capital letters written in the encoding block and the candidate list may indicate motion information of the corresponding encoding block, and different letters may indicate different motion information. The blurred motion information refers to the previously reconstructed motion information, and the motion information displayed in dark color may indicate newly generated motion information in the current encoding block.

[00332] In FIG. 14, step (a) shows a process in which the first coded block among the coding blocks included in the UPU is decoded. Since the pre-created motion information has not yet been created, there is no stored motion information in the candidate list. In step (b), the motion information A recovered in step (a) is added to the candidate list. Therefore, in predicting the motion information of the current encoding block, it is possible to refer to the motion information A stored in the candidate list.

[00333] Then, as the next steps progress, the motion information recovered in the previous step can be added to the candidate list one by one. Here, motion information overlapping with motion information already present in the candidate list cannot be added to the candidate list again. For example, the motion information generated in step (d) is already included in the candidate list as A and thus is not added. Therefore, in step (e), the newly added motion information may not exist in the candidate list. Similarly, the motion information generated in step (e) is already included in the candidate list as C and thus is not added. Therefore, the newly added motion information in the candidate list may not exist in step (f).

[00334] As another example, even if the motion information already exists in the candidate list, such as the motion information generated in steps (d) and (e), the motion information can be sequentially added to the candidate list. In this case, the list of candidates in step (e) could be A, B, C, A and C sequentially, starting at index 0.

[00335] FIG. 15 is another diagram illustrating a method in which motion information is added to a candidate list in accordance with the present invention.

[00336] FIG. 15 shows another example in which motion information is added to the candidate list. According to FIG. 15, motion information can be generated in the same manner as in FIG. 14, and motion information can be added to the candidate list at the same time. However, FIG. 15 illustrates a method of adding motion information to a candidate list in a manner different from the method of FIG. 14.

[00337] In FIG. 14, motion information is added to the last index position of the candidate list, but in the example of FIG. 15, the newly output motion information can be added to the index position 0 of the candidate list. In addition, motion information stored in the candidate list existing at index position 0 may change the order in the candidate list by adding one to the index when new motion information is added. In other words, by changing the order of the candidate list, one can change the order or priority of accessing the candidate list in motion prediction.

[00338] FIG. 16 is another diagram illustrating a method for adding motion information to a candidate list in accordance with one embodiment of the present invention.

[00339] FIG. 16 illustrates a method for reconfiguring the order of motion information in a candidate list when the same motion information as the newly inputted motion information exists among the previously input motion information in the method according to FIG. 15. The motion information A generated in step (d) is the same motion information as the motion information added to the candidate list in step (b). Therefore, in FIG. 16, in step (e), traffic information A cannot be added to the candidate list. In step (e) in FIG. 16, the order in the candidate list of motion information A may be changed to index 0. Alternatively, existing motion information A may be deleted, and motion information A may be newly added at index position 0. The index of other previously entered motion information is obtained by adding one, and the order and index between candidates can be changed.

[00340] This may be a method of changing the priority with new input information when the same motion information reappears taking into account the priority according to the order in which the motion information is input. Meanwhile, in step (f), it can be confirmed that when the order (index) of motion information in the candidate list is later than the overlapping motion information (when the index is larger), the order is not changed. That is, in the example in FIG. 16, when overlapping motion information C occurs, when the order of C changes to index 0, the index of A, which is lower than that of C, becomes +1, while the index of B, which is higher than that of C, may not change.

[00341] Meanwhile, at least one of an indicator (flag, index) indicating whether to use the candidate list in accordance with the present invention, the type of UPU used, the size of the UPU may be signaled by at least one of a set of parameters such as a set of parameters for a sequence or a set of parameters for a picture or slice title.

[00342] FIG. 17-20 are diagrams illustrating syntax structure in accordance with some embodiments of the present invention.

[00343] In the following description, the presented syntax is one example, and the scope of the present invention is not limited to the following syntax expression. Syntax having the same meaning or function as the syntax for a candidate list in accordance with the present invention may be included within the scope of the present invention.

[00344] FIG. 17 shows an example in which candidate list enable signaling is included in the sequence parameter set (SPS) syntax. Fig. 18 illustrates an example in which candidate list enable signaling is included in picture parameter set (PPS) syntax.

[00345] The candidate list enable signal may mean a signal indicating whether the candidate list in accordance with the present invention is used in encoding/decoding. In the tables in Fig. 17 and 18, amvol_enable_flag may be a candidate list enable signal indicating whether the candidate list is used in encoding/decoding. When amvol_enable_flag is true, encoding and decoding can be performed using the candidate list.

[00346] upu_type_idx may be a signal signaled when encoding and decoding are performed using the candidate list. upu_type_idx is a signal identifying the type of the UPU, and may be a signal identifying the unit to which the UPU applies. Here, the unit to which the UPU is applied may be at least one of a block, a CTU, a brick, a tile, a slice, a sub-picture, a picture, a frame, and a Group of Pictures (GOP) consisting of one or more CUs, and the UPU may have an area that coincides with the area of a specific unit.

[00347] In this case, encoding can be performed by assigning a unit to which the UPU applies and assigning a specific value to it. UPU_TYPE_COSTOM may be syntax indicating that the UPU unit is not a predefined unit among the units to which the UPU applies. When the unit to which the UPU applies is UPU_TYPE_COSTOM, additional signaling may be performed to set the range of the UPU. For example, the upu_width and upu_height signals in FIG. 17 and 18 could be syntax for signaling the size to which the UPU is applied, in terms of width and height.

[00348] upu_type_idx may be omitted when the UPU generates a fixed-size candidate list or the UPU type may be inferred from another signal. When a UPU is applied only in a unit or size specified in advance, the syntax signaling for setting UPU_TYPE_COSTOM and the UPU range can be omitted.

[00349] FIG. 19 illustrates a case in which candidate list enable signaling is included in the slice header syntax. amvol_enable_flag may be a signal indicating whether the candidate list in accordance with the present invention is used for encoding/decoding. When amvol_enable_flag is true, encoding and decoding can be performed using the candidate list in accordance with the present invention.

[00350] Upu_type_idx may be a signal that may be transmitted and received when performing encoding and decoding using the candidate list. upu_type_idx is a signal identifying the type of the UPU, and may be a signal identifying the unit to which the UPU applies. Here, the unit to which the UPU is applied may be at least one of a block, a CTU, a brick, a tile, a slice, a sub-picture, a picture, a frame, and a Group of Pictures (GOP) consisting of one or more CUs, and the UPU may have an area which coincides with the area of a certain unit.

[00351] In this case, encoding can be accomplished by assigning a unit to which the UPU applies and assigning a specific value to it. UPU_TYPE_COSTOM may be syntax indicating that the UPU unit is not a predefined unit among the units to which the UPU applies. When the unit to which the UPU applies is UPU_TYPE_COSTOM, additional signaling may be performed to set the range of the UPU. For example, the upu_width and upu_height signals in FIG. 17 and 18 could be syntax for signaling the size to which the UPU is applied, in terms of width and height.

[00352] upu_type_idx may be omitted when the UPU generates a fixed-size candidate list or the UPU type may be inferred from another signal. When a UPU is applied only in a unit or size specified in advance, the syntax signaling for setting UPU_TYPE_COSTOM and the UPU range may be omitted.

[00353] Here, all signaling regarding the candidate list can be omitted when the slice type is a concrete type. For example, all signaling related to the candidate list may be omitted when the current slice type is I-slice. Since the candidate list is an encoding and decoding method used for motion prediction, signaling can be omitted in the case of an I-slice that does not perform motion prediction.

[00354] FIG. 20 shows an example where the candidate list index is included in the encoding unit syntax. The candidate list index signal according to the present invention may be a syntax indicating which candidate motion information from the candidate list candidates is used to predict motion information when motion information prediction using the candidate list is used.

[00355] As an example, the candidate list index signal can be used in combination with a merge mode, an AMVP mode, an IBC mode, etc. and can be used in conjunction with other signals without any separate signaling. Alternatively, the signaling may be omitted using the candidate's previously assigned motion information. As another example, the candidate list index signal may be omitted when the current coding block is not in inter-prediction mode.

[00356] In FIG. 20, cu_inter_pred_mode may be a signal indicating which encoding method the current block is encoded/decoded when the current CU is in inter-prediction mode. When cu_inter_pred_mode indicates an encoding/decoding method using a candidate list according to the present invention, a candidate list index signal may be signaled. As another example, the cu_inter_pred_mode signaling may be omitted and the encoding/decoding method of the current encoding block may be selected in a predetermined manner. The encoding/decoding method of the current encoding block is selected in a predetermined manner, and when the predetermined encoding/decoding method is a candidate list encoding/decoding method according to the present invention, a candidate list index signal may be signaled.

[00357] As an example, MODE_AMVOL may be a syntax indicating an encoding/decoding method using a candidate list in accordance with the present invention. amvol_idx can indicate the candidate list index. By the signaled amvol_idx, it is possible to determine which motion information among the motion information stored in the candidate list is used as a candidate for motion prediction.

[00358]

[00359] Next, a method for approximating motion information included in a candidate list according to the present invention will be described.

[00360] The motion information stored in the candidate list according to the present invention may be similar to each other. When similar motion information exists in a large number of candidate lists, a problem may arise that the characteristics of the motion information that is a candidate are shifted so that the prediction performance of the motion information is reduced. Here, the above problem can be solved by transforming and using similar candidates in one approximated motion information.

[00361] The time point when the approximate motion information is generated is one of the time point when the motion information is stored in the candidate list according to the present invention, the time point when the candidate list candidate is generated from the candidate list, and the time point when the motion information , stored in the list of candidates for each period, is approximated.

[00362] FIG. 21 and 22 are diagrams illustrating flowcharts of an encoder and decoder that approximate motion information in accordance with some embodiments of the present invention.

[00363] FIG. 21 and 22 illustrate flowcharts of an encoder/decoder using a candidate list motion information approximation method in accordance with the present invention. Fig. 21 illustrates a case where motion information of a candidate list is approximated at the time of adding motion information to the candidate list, and FIG. 22 illustrates a case where candidate list motion information is approximated during candidate list candidate generation.

[00364] FIG. 21 and 22 may be examples in which a motion information approximation step is added to the flowchart of FIG. 11. The encoder or decoder may approximate the motion information while adding the motion information to the candidate list. The motion information approximated in the motion information approximation step may be added again to the candidate list in the candidate list motion information adding step.

[00365] In FIG. 22, when an encoder or decoder generates a candidate list candidate from motion information in the candidate list, the encoder or decoder can approximate the motion information of a candidate list candidate. The candidate list motion information prediction step may be performed by candidates of the candidate list having approximated motion information. Moreover, since each step in FIG. 21 and 22 may be the same as each step in FIG. 11, its description will be omitted.

[00366] FIG. 23 and 24 are diagrams illustrating the structure of an encoder and decoder for approximating motion information in accordance with some embodiments of the present invention.

[00367] FIG. 23 and 24 show example circuits of sub-decoder devices using a candidate list motion information approximation method. Fig. 23 illustrates a case where motion information of the candidate list is approximated at the time of adding motion information to the candidate list, and FIG. 24 illustrates a case where candidate list motion information is approximated during candidate list candidate generation.

[00368] FIG. 23 and 24 illustrate examples in which a candidate list motion information approximation unit is added to the apparatus circuit of FIG. 12. The candidate list motion information approximation unit can approximate motion information when motion information is added to the candidate list. The motion information approximated by the candidate list motion information approximation unit may be added again to the candidate list by the candidate list motion information appending unit.

[00369] The candidate list motion information approximation module according to FIG. 24 can approximate movement information of candidates in the candidate list when generating a candidate list candidate from the movement information in the candidate list. The candidate list motion information prediction unit may perform motion information prediction using the candidate list candidates having approximated motion information. Moreover, since each component in FIG. 23 and 24 may be the same as each component in FIG. 12, its description will be omitted.

[00370] Next, a method for approximating motion information by an encoder or decoder in accordance with some embodiments of the present invention will be described.

[00371] As an example, the motion information can be approximated using the average value of the motion information. An encoder or decoder may approximate motion information corresponding to a certain range of similarity by obtaining an average value of motion information corresponding to a certain range of similarity. Here, at least one of the arithmetic average, the weighted average, and the median may be used as a method for calculating the average. In addition, the average value of the motion information can be calculated within a range determined regardless of the similarity.

[00372] As another example, motion information can be approximated by representative selection on the frequency of occurrence of motion information. After obtaining the same number of pieces of motion information within a predetermined range, the encoder or decoder can approximate the same motion information within the predetermined range by defining the largest number of pieces of motion information as representative motion information. For example, when the predetermined range is a CTU, the encoder or decoder may compare all the motion information generated in one CTU and determine the motion information with the most occurrences as representative motion information of the current CTU. The representative motion information can be used as approximate motion information.

[00373] In addition, the above-described determination of motion representative information can be performed using similar motion information instead of the same motion information. In this case, similar motion information may mean motion information having a difference within a particular threshold value with respect to the reference motion information. The encoder or decoder can confirm the number of occurrences by treating similar motion information as the same motion information described above, and determine representative motion information based thereon.

[00374] As another example, motion information may be approximated taking into account a portion of the region from which motion information is derived. The portion of the area from which motion information is output may be different for each piece of motion information. The encoder or decoder may approximate the motion information based on a portion of the region from which the motion information is output. For example, motion information is output from a large area, the encoder or decoder may calculate a weighted average by giving a higher weight, or approximate the motion information by selecting motion information output from the largest area as representative motion information. In this case, when motion information obtained from regions different from each other has the same motion with respect to each other, the regions can be summed and taken into account.

[00375] As another example, motion information may be approximated taking into account the prediction potential of motion information. The encoder or decoder may compare pieces of prediction potential information with motion information to be added to the candidate list to select motion information having high prediction potential as motion representative information, or calculate motion representative information using a weighted average according to with predictive potential. Here, the encoder or decoder may set a range for comparing the prediction potential or approximate the motion information by comparing the prediction potential only between similar pieces of motion information.

[00376] Here, the motion information prediction potential may be information indicating how high the motion information prediction accuracy or coding efficiency is when motion information prediction using the candidate list is used in accordance with the present invention. A method for determining motion information prediction potential will be described below.

[00377] As another example, motion information can be approximated by setting the resolution of the motion information to a specific pixel unit. The motion vector to be added to the candidate list can be approximated to a specific pixel unit. For example, a particular pixel may be one of an integer pixel unit, a 1/2-pixel unit, a 1/4-pixel unit, and a four-pixel unit. For example, when the resolution of a motion vector added to the candidate list is approximated to 1/16-pixel unit and integer pixel unit, the value of the motion vector can be approximated to the nearest integer value. As another example, when approximating to a 1/2-pixel unit, the motion vector value may be approximated to the nearest 1/2-unit value. For example, when the motion vector value is (2.6; 3.7), when approximating to the nearest 1/2 unit value, the motion vector value can be approximated to (2.5; 3.5).

[00378] Approximation of motion information can be performed using the above-described embodiments separately. In addition, approximation of motion information can be performed by at least one combination of the above-described embodiments.

[00379]

[00380] Next, a method for performing prediction using the candidate list motion information prediction potential according to the present invention will be described.

[00381] The motion information stored in the candidate list in accordance with the present invention may include additional information for more efficient use of the candidate list in addition to the motion prediction information. The prediction potential of motion information included in the candidate list may be determined using some or all of the motion information and additional information included in the candidate list.

[00382] In the present description, the prediction potential may be information indicating how high the prediction accuracy or encoding efficiency of motion information is when used to predict motion information using a candidate list. Therefore, it is possible to improve encoding efficiency by first selecting motion information having high prediction potential and generating candidates for the candidate list.

[00383] A method for determining a motion information prediction potential according to the present invention may include a method for calculating a global potential of motion information belonging to a UPU candidate list and a method for calculating a local potential for a block (unit) to be encoded at the current time.

[00384] The method for calculating the global potential may be a method for estimating how promising each piece of motion information included in the candidate list is for the entire UPU. The method for calculating the local potential may be a method for assessing how promising each piece of motion information included in the candidate list is for predicting the motion information of a particular block.

[00385] FIG. 25 and 26 are diagrams illustrating a method for calculating candidate list potential in accordance with some embodiments of the present invention.

[00386] FIG. 25 and 26 show flowcharts of an encoder and a decoder that calculate motion information prediction potential. Fig. 25 illustrates a case where the motion information prediction potential is calculated at the time of adding motion information to the candidate list, and FIG. 26 illustrates a case where the prediction potential of motion information included in the candidate list is calculated during candidate generation of the candidate list.

[00387] FIG. 25 and 26 may be examples in which a prediction potential calculation step is added to the flowchart of FIG. 11. For example, calculating the candidate list prediction potential according to FIG. 25 can be performed by the method described above for calculating the global potential. The motion information added to the candidate list may include prediction potential information. When the same motion information exists in the existing candidate list, the prediction potential information can be updated.

[00388] For example, the prediction potential calculation step in FIG. 26 can be performed using the local potential calculation method described above.

[00389] Steps for predicting candidate list motion information in FIG. 25 and 26 may be performed in consideration of the output prediction potential information. Moreover, since each step in FIG. 25 and 26 may be the same as each step in FIG. 11, its description will be omitted.

[00390] FIG. 27 and 28 are diagrams illustrating the structure of an encoder and decoder for computing candidate list prediction potential in accordance with some embodiments of the present invention.

[00391] FIG. 27 and 28 show example circuits of encoder and decoder devices that calculate motion information prediction potential. Fig. 27 illustrates a case where the motion information potential is calculated at the time of adding motion information to the candidate list, and FIG. 28 illustrates a case where the candidate list motion information potential is calculated during candidate list candidate generation.

[00392] FIG. 27 and 28 show examples in which a prediction potential calculator is added to the device circuit of FIG. 12. For example, the prediction potential calculator according to FIG. 27 calculates the potential of motion information to be added to the candidate list, and can calculate the prediction potential using the global potential calculation method described above. The motion information added to the candidate list may include prediction potential information. When the same traffic information exists in an existing candidate list, the candidate information can be updated.

[00393] Meanwhile, the prediction potential calculator according to FIG. 28 calculates the prediction potential when generating a candidate list candidate, and the prediction potential can be calculated using the local prediction potential calculation method described above.

[00394] The candidate list motion information prediction modules of FIG. 27 and 28 can predict motion information taking into account the output prediction potential information. Moreover, since each component in FIG. 27 and fig. 28 may be the same as each component in FIG. 12, its description omitted.

[00395] FIG. 29 is a diagram illustrating the generation of a mixed candidate list in accordance with an embodiment of the present invention.

[00396] FIG. 29 shows an example circuit of an encoder and decoder apparatus for calculating a motion information prediction potential using a global potential calculation method and a local potential calculation method. Device diagram in Fig. 29 can be obtained by replacing the prediction potential calculator according to FIG. 27 with a global potential calculator and replacing the prediction potential calculator according to FIG. 28 local potential calculator. According to FIG. 29, candidate list candidates can be generated taking into account both the global potential and the local potential, which can be used in the candidate list motion information prediction unit.

[00397] Next, a method for calculating motion information prediction potential in accordance with some embodiments of the present invention will be described.

[00398] As an example, the prediction potential can be calculated taking into account the frequency of occurrence of motion information. The motion information prediction potential can be calculated by checking how much motion information in the candidate list has already been generated. When the frequency of occurrence of the same motion information or similar motion information that can be treated as the same motion information is high, the encoder or decoder may determine that the prediction potential of the corresponding motion information is high. Here, the frequency of occurrence can be measured using the number of occurrences of the same motion information, an area having the same motion information, and the like. Here, the occurrence frequency can be measured by limiting the area where the occurrence frequency is measured. For example, the occurrence rate to date may be measured for the entire UPU region, or the occurrence rate may be measured only within a predefined range or specific region based on the unit that is being encoded at the current time. The frequency of occurrence measured in this way can be used to calculate the prediction potential of motion information.

[00399] As another example, the prediction potential may be calculated taking into account inter-prediction accuracy. Since each piece of motion information belonging to the candidate list is motion information used to reconstruct another block, the encoder or decoder can know the coding efficiency of a block reconstructed using the corresponding motion information.

[00400] The efficiency of encoding motion information can be determined based on the energy distribution or the average value of the residual signal. The residual signal energy may indicate a difference in brightness value or a difference in pixel value between the current block and the prediction block generated by inter-prediction. The average residual signal energy value may mean the average value of the difference between a luminance value or a pixel value generated at each pixel position of the current block and the prediction block. The residual signal energy distribution may be a value calculated through the residual signal energy position for each pixel position in the differential block and the energy amount. For example, the residual signal energy distribution may be the torque value of residual signals having an energy of a predetermined amount or more. Here, the moment value of the residual signals may denote the central position of the distribution of luminance values, which is calculated by the positions of the corresponding residual signals and the luminance values of the corresponding residual signals.

[00401] From the energy of the residual signal, a distribution and an average value and the like can be calculated. in a frequency-converted differential block. For example, the DC value can be used as an average in the frequency domain, and the energy distribution of the residual signals can be determined through the energy ratio for each frequency domain. Here, to make the encoder and decoder work identically, a quantized differential block or a frequency-converted difference block can be used. In addition, when using a quantized differential block or a frequency-converted differential block, a quantization parameter or quantization step value may be used. For example, when comparing the energy distribution or average value of the residual signal of two blocks using different quantization parameters from each other, the energy distribution or average value of the differential signal can be corrected by the difference of the quantization parameter.

[00402] For example, after the energy of a differential signal having a larger quantization parameter is scaled n times, the energy distribution or average of the differential signal can be calculated to compare the energy of the residual signals between the two blocks. In addition, since the high-frequency component is preferably reduced during quantization, when comparing the energy of the residual signal in the frequency domain, the calculation can be performed by assigning separate weights to the low-frequency component and the high-frequency component.

[00403] Here, when the average energy value of the residual signal is high, it is determined that the prediction potential is low. The prediction potential can be determined by a method of comparing the energy distribution of a residual signal by changing its priority according to the position of the residual signal energy distribution, and a method of comparing the energy distribution of the residual signal according to a frequency region such as a high-frequency region and a low-frequency region.

[00404] That is, when the encoding efficiency is high when using the corresponding motion information, by comparing the encoding efficiency of the reconstructed blocks using each piece of motion information, the encoder or decoder can determine that the corresponding motion information has a high prediction potential.

[00405] As another example, the prediction potential can be calculated taking into account the distance between the position of the current block and the position of the region in which motion information has appeared. Motion information inferred from a specific region is more likely to be similar to a specific region that is temporally and spatially adjacent to the current block. Considering such a characteristic, the prediction potential can be determined according to the spatial and temporal distance between the generation position of each piece of motion information included in the candidate list and the position of the current block.

[00406] Here, in the case of the same motion information included in the candidate list, the encoder or decoder can reduce the operations required to calculate the distance by using approximated position information or a method for updating the most recent position. As an example of a method for calculating approximate position information when motion information is generated in units of encoding blocks, a method for simplifying distance calculation taking into account only the position information of a CTU including a corresponding encoding block may be used.

[00407] When the candidate list according to the present invention is used in combination with a fusion mode, an AMVP mode, an IBC mode or the like, the prediction potential can be determined excluding the adjacent region determination method, which is taken into account by the existing motion information prediction method.

[00408] Calculation of motion information prediction potential can be performed using the above-described embodiments separately. In addition, calculation of the motion information prediction potential can be performed by at least one combination of the above-described embodiments. When the above-described embodiments are combined to calculate the prediction potential, the prediction potential can be calculated by weighing each determination result value and then applying the importance of each determination result differently.

[00409]

[00410] Next, a method for generating a virtual candidate list in accordance with an embodiment of the present invention will be described.

[00411] Candidate list candidates in accordance with the present invention can be generated using motion information previously stored in the candidate list. For example, candidate list candidates may be generated using all or some of the motion information previously stored in the candidate list. Further, a candidate having motion information that is derived from the motion information previously stored in the candidate list, but not identical to the motion information previously stored in the candidate list, is defined as a virtual candidate. According to an embodiment of the present invention, a virtual candidate may be added to a candidate list.

[00412] Next, a method for generating a virtual candidate will be described.

[00413] For example, a virtual candidate may refer to a candidate from a mixed list of candidates. For example, there may be a case where the prediction is performed using multiple pieces of motion information. In particular, prediction using two pieces of motion information may be called bidirectional prediction.

[00414] The motion information used for inter-prediction using multiple pieces of motion information may be configured as a set of multiple pieces of motion information. Candidates in a candidate list having a set of multiple pieces of motion information as motion information can be generated from motion information stored in different candidate lists for each piece of motion information constituting the set. When the motion information stored in the candidate list is a set of multiple pieces of motion information, candidates of the mixed candidate list can be generated using some or all of the motion information.

[00415] FIG. 30 is a diagram illustrating the generation of a mixed candidate list in accordance with an embodiment of the present invention.

[00416] FIG. 30 shows an example in which motion information of a mixed candidate list is generated. Example in Fig. 30 shows a case where the motion information of the mixed candidate list has motion information for reference picture lists L0 and L1, and each piece of motion information consists of a motion vector and a reference picture number (index, identifier, number, etc.). In the example in FIG. 30, the mixed candidate list has two sets of motion information A and B, and each set of motion information includes motion information from the reference picture list L0 and motion information from L1.

[00417] Here, C1 and C2 may mean mixed candidate list motion information generated from motion information sets A and B. Here, C1 may be generated from motion information A from L0 and motion information B from L1, and C2 may be generated from motion information B from L0 and motion information A from L1.

[00418] As another example, candidates of a new candidate list may be generated by an average or weighted average of some or all of the motion information previously stored in the candidate list. The virtual candidates obtained in this way can be defined as the average candidate.

[00419] For example, the middle candidate may be derived from two or more pieces of motion information output in a predetermined manner, motion information outside the candidate list not stored in the candidate list motion information, predetermined motion information, and the like may be used. in the process of deducing motion information. The average candidate generation process may use the average of all motion information and may use the average of some motion information. For example, when the motion information consists of a motion vector and a reference picture number, the average of the motion vector and the reference picture number can be calculated to generate an average candidate, and the average of one of the motion vector and the reference picture number is calculated to generate an average candidate.

[00420] The encoder or decoder may add the derived average candidate to the candidate list. For example, when a middle candidate is added, the largest index value may be assigned to the candidate list. There may be different ways to configure the weight. For example, the weight may be set according to the prediction potential.

[00421] Generation of a virtual candidate can be performed by using the above-described embodiments separately. In addition, generation of virtual candidates may be performed by a combination of at least one of the above-described embodiments.

[00422]

[00423] Next, a method for initializing movement information of a candidate list in accordance with the present invention will be described.

[00424] A list of candidates in accordance with the present invention may be generated in the UPU described above. Here, the UPU may be one of a block, CTU, brick, tile, slice, sub-picture, picture, frame and group of pictures (GOP) consisting of one or more CUs. In the UPU, the candidate list is again created, used and initialized. Next, initialization of a candidate list in accordance with some embodiments of the present invention will be described.

[00425] For example, a candidate list in accordance with the present invention may be newly created and used for each UPU. That is, a candidate list may be initialized for each UPU. The candidate list may not have motion candidates saved immediately after initialization. A candidate list may be generated for each UPU, and there may be no correlation between each UPU's candidate lists. For example, a candidate list may be initialized for each CTU.

[00426] According to the initialization method, it is possible to prevent heterogeneous motion information from appearing as a candidate and re-perform encoding from the starting point of the UPU without the previous prediction information.

[00427] As another example, the current UPU candidate list may be initialized using UPU motion information that is already encoded, or the UPU candidate list. Here, a UPU to be referenced may be specified, and the candidate list of the current UPU may be initialized using the traffic information or candidate list of the corresponding UPU.

[00428] Here, indicating the UPU to be referenced may be performed by a method of allowing the encoder and decoder to refer to the same UPU by a signaled signal, and a method of skipping transmission/reception of an additional signal by selecting the UPU in a predetermined manner.

[00429] Initialization of the current candidate list can be accomplished by copying all or part of the UPU candidate list to which it refers. Alternatively, the current candidate list may be initialized based on the prediction potential or priority of motion information that is output or copied.

[00430] FIG. 31 and 32 are diagrams illustrating a method for initializing a candidate list in accordance with some embodiments of the present invention.

[00431] FIG. 31 shows an example of initializing the candidate list of the current UPU using the candidate list of the restored UPU when the UPU is a picture unit. According to FIG. 31, motion information A', B' and C' are added to the candidate list before recovering the first motion information of the current picture, which is copied from the candidate list of the reference picture, which is the reconstructed UPU.

[00432] FIG. 31 shows the case of copying part of the candidate list of the restored UPU. In fig. 31, the candidate list index indicates the prediction potential or priority of a candidate list candidate, which indicates that the smaller the number, the higher the prediction potential or the higher the priority. According to FIG. 31, the current frame can be restored by configuring the priority of the copied motion information. Here, the priority of the previously copied motion information A', B' and C' can be set to have a higher priority than the newly added motion information A and B.

[00433] FIG. 32 shows an example of copying motion information to initialize a candidate list in both the L0 direction and the L1 direction. As shown in FIG. 31, the recovered UPU can be used to initialize the candidate list, and the UPU can be installed in the picture unit. According to FIG. 32, the current candidate list may be initialized by copying motion information from the candidate lists of pictures used for support in the L0 direction and the L1 direction. For example, three motion information A', B' and C' can be copied from list L0, and two motion information A'', B'' can be copied from list L1. As described above, when copying motion information used to initialize a list of candidates from UPUs different from each other, the number of pieces of motion information referenced from each recovered UPU can be set differently.

[00434] Information about whether initialization of the candidate list is performed, picture information referenced for initialization, number of pieces of motion information referenced for initialization, etc. may be signaled at a higher bitstream level, such as a sequence parameter set, a picture parameter set, a slice header, etc.

[00435]

[00436] Next, a method for scaling candidate list motion information in accordance with the present invention will be described.

[00437] When performing inter-prediction, prediction may be performed on the current picture using a temporally adjacent reference picture. The current picture and the reference picture can exist in different positions on the time axis. When inter-prediction is performed, motion information can be output taking into account the motion of an object or background in a picture due to the time difference between the current picture and the reference picture.

[00438] Here, the reference picture referenced by the current block and the reference picture referenced by the candidate block may be different from each other. Therefore, when deriving motion information using a candidate block, it is necessary to adjust the motion information of the current block taking into account the temporal distance between the pictures to which the referencing block and the referenced block belong.

[00439] By comparing the temporal distance between the picture to which the current block belongs and the picture referred to by the current block, and the temporal distance between the picture to which the candidate block belongs and the picture referred to by the candidate block, a method for correcting motion information candidate block to be more similar to the motion information of the current block is called motion information scaling. Motion information scaling can be used in a candidate inter-prediction method such as AMVP mode, fusion mode, and the like.

[00440] FIG. 33 is a diagram illustrating scaling of motion information.

[00441] FIG. 33 shows an example of scaling motion information. t may denote a position on the time axis of the current picture, and t-n may denote a position on the time axis of the reference picture of a candidate block. t-m can mean the position on the time axis of the reference picture of the current block. When the position on the time axis of each picture changes, the movement of the image within the picture can be determined. A can mean the current block, and B can mean a candidate block. A' may denote a position in the reference picture of a current block having the same spatial position as the current block A. B' may denote a position in the reference picture of a candidate block having the same spatial position as candidate block B.

[00442] Since candidate block B performs inter-prediction from the reference picture of the candidate block, it is possible to derive the position of block D as a result of inter-prediction, and express this position as motion vector MV1. When deriving the motion vector of the current block A from the candidate block B, the time distance between the current picture and the reference picture of block A (for example, t-(t-m)=m) and the time distance between the current picture and the reference picture of block B (for example, t- (t-n)=n) are different from each other, motion information scaling can be applied.

[00443] For example, in FIG. 33, using the motion vector MV1 of block B, the temporal distance m between the current picture and the reference picture of block A, and the temporal distance n between the current picture and the reference picture of block B, the motion vector (MV1*m/n) of the current block A can be derived. Since the vector motion (MV1*m/n) to which the motion information scaling is applied is similar to the motion vector MV0 of the current block more than MV1, the motion information prediction efficiency can be increased.

[00444] Motion information scaling may be a method of scaling motion information using the temporal distance between pictures, assuming that the moving object in each picture of the video has linear motion.

[00445] Since the current block A performs inter-prediction from the reference picture of the current block, the position of block C can be output, and this position can be expressed as a motion vector MV0. However, since the pictures referenced by blocks A and B are different from each other, the region suitable for predicting the current block in each reference picture may be different from each other. Therefore, a size difference appears between MV0 and MV1.

[00446] Here, motion information scaling may be applied using the distance between the current picture and the reference picture of the candidate block on the time axis and the distance between the current picture and the reference picture of the current block on the time axis. The corrected motion information is more likely to be similar to the motion information of the current block than the uncorrected motion information. It is possible to improve coding efficiency by using more similar motion information for prediction.

[00447] That is, when motion vector scaling is applied to MV1 in FIG. 33, a motion vector modified close to MV0 can be output. An encoder or decoder can improve coding efficiency using a scaled motion vector.

[00448] Since the candidates of the candidate list in accordance with the present invention may also have reference pictures other than the current block, it is necessary to perform scaling of the motion information taking into account the time interval between the current block and the reference picture and the time interval between the candidates of the candidate list and the candidate reference picture. For example, when candidate list candidates use motion information of temporally adjacent blocks, the motion information scaling described above may be required. Therefore, information for scaling the motion information can be stored as additional information in the motion information stored in the candidate list.

[00449] For example, the information for scaling motion information may include at least one of a POC number of a reference picture, a POC number of a picture in which motion information appears, and time distance information between a picture in which motion information appears and the reference picture .

[00450] When the UPU is a unit including a plurality of pictures, such as a GOP, the motion information stored in the candidate may relate to reference pictures different from each other. When motion information is predicted using motion information stored in the candidate list, the motion information reference picture included in the referenced candidate list and the reference picture of the current block may be different from each other. In addition, when the UPU includes a plurality of pictures, the pictures in which motion information stored in the candidate list may appear may be different from each other.

[00451] For example, traffic information A referring to POC 2 in picture POC1 may be stored in a candidate list, and motion information B referring to POC 4 in picture POC 3 may be stored in a candidate list. In this case, the picture in which motion information A stored in the candidate list appears may be POC 1, and the picture in which motion information B stored in AMVOL appears may be POC 3. The picture referenced by motion information A is may be POC 2 and the picture referenced by motion information B may be POC 4.

[00452] In the case of a UPU including a plurality of pictures, the picture in which motion information stored in the candidate list appears and the reference picture may be different from each other. In addition, the picture in which the motion information stored in the candidate list appears and the current picture including the current block may also be different from each other, and the reference picture of the motion information stored in the candidate list and the reference picture of the current block also may differ from each other. Here, when the motion information of the current block is predicted through the motion information of the candidate list, it is possible to improve the encoding efficiency by scaling the motion information as described above.

[00453] FIG. 34 is a diagram for explaining the scaling of motion information according to an embodiment of the present invention.

[00454] For example, when the UPU is a picture unit, scaling may be performed as shown in FIG. 34. In the example of FIG. 34, UPU A may store motion information generated in the POC picture 8 in a candidate list, and UPU B may store motion information generated in the POC picture 4 in a candidate list. Picture POC 8 can perform motion prediction by referring to picture POC 1, and picture POC 4 may perform motion prediction by referring to picture POC 8. When picture POC 4 is the current picture to be encoded at the current time, POC 8 is the picture which has already been restored, the candidate list of UPU B can be initialized by accessing the candidate list used in UPU A.

[00455] However, the motion information stored in the candidate list of UPU A is motion information generated with reference to picture POC 8, and thus differs in the time distance between pictures in the case where picture POC 4 refers to picture POC 8. Meanwhile, since the picture to be encoded with reference to the motion information stored in the candidate list of UPU B is POC 4, and the picture referenced by the picture of POC 4 is POC 8, the motion information of the candidate list of UPU A is used as is, so that in the case of initializing the UPU B candidate list, the motion information prediction performance is reduced. Here, when the motion information of UPU A is adjusted by scaling the motion information, and the scaled motion information is used to initialize the candidate list of UPU B, the coding efficiency can be improved.

[00456]

[00457] Next, a method for sharing a candidate list in accordance with the present invention will be described.

[00458] When performing encoding and decoding, the candidate list information may vary according to the time point at which each encoding block performing inter-prediction is referenced. Since candidate list movement information may be added, deleted, or changed, information or configuration included in the candidate list may vary according to the point in time when inter-prediction is performed. That is, when the time at which each inter-prediction region accesses the candidate list varies, the reference information may change. Since the candidate list according to the present invention can be updated as a result of performing inter-prediction, inter-prediction must be performed sequentially.

[00459] In the following description, an area consisting of one or more coding blocks may be defined as a common (shared) area. When inter-prediction is performed, encoding blocks included in the common area may share a candidate list configured at a certain time earlier than the encoding or decoding time of all encoding blocks included in the common area. Here, the common area may be configured by at least one unit or a set of CTUs, CUs and PUs.

[00460] Accordingly, inter-coding may be performed independently or in parallel for each encoding block in the corresponding common area. When inter-prediction is performed by accessing the candidate list independently or in parallel in the common area, each encoding unit included in the common area and performing inter-prediction may access the candidate list that was previously configured. Here, a preconfigured candidate list may mean a candidate list configured before encoding or decoding begins.

[00461] Here, encoding blocks in a common area may share a preconfigured candidate list or may share candidates from a candidate list generated using the candidate list. A candidate list candidate suitable for each encoding block can be separately generated by accessing the preconfigured candidate list during inter-prediction of encoding blocks included in the common area. That is, when the present method is applied, a preconfigured candidate list accessed by coding blocks included in the common area can be shared, and candidates from the candidate list can be generated independently. Alternatively, coding blocks included in the common area may share not only a preconfigured candidate list, but also candidates of candidate lists derived from the common candidate list.

[00462] After prediction is performed in encoding blocks included in the corresponding common area, the candidate list can be updated independently or in parallel using motion information or prediction information generated in each of the encoding blocks. In this case, the list of candidates can be updated in a predetermined order after all predictions in all coding blocks included in the common area are completed. Accordingly, the encoder and decoder can generate the same list of candidates when encoding/decoding the current block.

[00463] For example, when the current block is encoded/decoded in the triangular partitioning mode, the current block may be divided into two triangular areas. Here, partitioning direction information for dividing the current block into two triangular areas can be encoded/decoded. To encode/decode each of two regions divided into triangular regions, motion information for encoding/decoding each region can be selected from a list of candidates. Here, two triangular regions obtained by performing triangular partitioning can share a block candidate list before partitioning.

[00464] When the current block is encoded/decoded in the triangular partition mode, the first motion information and the second motion information can be determined in the general candidate list through the first index and the second index for the current block. The encoder or decoder may output a first prediction block for the current block using the first motion information, and a second prediction block for the current block using the second motion information.

[00465] An encoder or decoder may generate a final prediction block for the current block by weighing the first prediction block and the second prediction block. Here, the weighted sum of the prediction blocks can be obtained by weighting the first region with respect to the first prediction block and weighting the second region with respect to the second prediction block.

[00466]

[00467] FIG. 35-39 are diagrams illustrating a method for updating and sharing a candidate list in accordance with some embodiments of the present invention.

[00468] FIG. 35 shows an example in which the candidate list is updated over time and an example of the candidate list accessed by a coding unit. Fig. 35 illustrates a case where encoding block 0 is encoded/decoded, and then encoding block 1 is encoded/decoded. The encoding block shown in gray indicates the encoding block that is currently being encoded/decoded. Encoding block 0 may be encoded/decoded with reference to a candidate list generated prior to the time at which encoding block 0 is encoded/decoded.

[00469] Thereafter, the candidate list can be updated with the movement information of encoded/decoded encoding block 0. The updated candidate list can be accessed when encoding/decoding encoding block 1, which is the next encoding block. The motion information of the recovered coding block 1 can be used to re-update the candidate list. The updated list of candidates can then be accessed again in the next coding block. Therefore, in order to encode/decode encoding block 1, encoding block 0 must first be encoded/decoded.

[00470] FIG. 36 illustrates a case in which the candidate list is updated over time, and a method for accessing the candidate list in each block when encoding block 0 and encoding block 1 constitute one common area. Since encoding block 0 and encoding block 1 are included in one common area, an encoder or decoder can perform prediction using a common candidate list for encoding block 0 and encoding block 1. Accordingly, unlike the example in FIG. 35, the encoder or decoder may perform inter-prediction on encoding block 0 and encoding block 1 independently or in parallel.

[00471] FIG. 36 shows a state in which prediction is performed in parallel using the same candidate list that was previously reconstructed at the same time. Unlike the example in FIG. 35, since the previously reconstructed candidate list is shared, encoding block 1 can access the candidate list at the time required for encoding/decoding even if encoding block 0 is not encoding/decoding.

[00472] In the example of FIG. 36, since both encoding block 0 and encoding block 1 are encoded/decoded, it is shown that motion information “E” generated from encoding block 0 and motion information “F” generated from encoding block 1 are reflected (updated) in list of candidates. Here, the candidate list may be updated at a time when encoding/decoding of the common area is completed. Fig. 36 illustrates a case in which the size of the candidate list is 5, and shows an example in which one motion information A from the candidate list is deleted in accordance with updating the motion information.

[00473] FIG. 37 shows an example of updating the candidate list and encoding/decoding encoding blocks 0 and 1 using the updated candidate list over time. A candidate from the candidate list suitable for the current encoding block may be generated from the updated latest candidate list, and the encoding block may be encoded/decoded with reference to the candidate from the generated candidate list. The motion information of the newly encoded/decoded encoding block is used to update the candidate list.

[00474] In the example of FIG. 37, encoding block 0 may be encoded/decoded earlier than encoding block 1. Accordingly, candidates for candidate lists corresponding to encoding block 0 can be generated using the updated candidate list before encoding block 0 is encoded/decoded, and the block encoding 0 can be encoded/decoded based on the candidate list. The motion information encoded/decoded in encoding block 0 can be used to update the candidate list, and the candidate list for encoding block 1 can be generated using the updated candidate list. The encoding block 1 may be encoded/decoded using a candidate from the candidate list generated for the encoding block 1. That is, according to FIG. 37, encoding blocks 0 and 1 can be encoded/decoded sequentially.

[00475] FIG. 38 illustrates an example of updating the common candidate list over time and encoding/decoding a coding block in each common region using the updated common candidate list. In this example, the common area may consist of encoding block 0 and encoding block 1. Encoding block 0 and encoding block 1 can generate candidates using the candidate list updated before encoding/decoding the two encoding blocks in the common area. Accordingly, the encoder or decoder may perform prediction independently or in parallel with respect to coding block 0 and coding block 1.

[00476] Unlike the example in FIG. 37, since in FIG. 38, a common candidate list is used, even if encoding block 0 is not encoded/decoded first, encoding block 1 can be encoded/decoded together with encoding block 0. That is, according to the example in FIG. 38, encoding block 0 and encoding block 1 can be encoded/decoded in parallel at the same time.

[00477] Referring to FIG. 38(a), encoding block 0 and encoding block 1 use a common candidate list, but may be encoded/decoded using candidates from different candidate lists. With reference to FIG. 38(b), encoding block 0 and encoding block 1 can generate candidates from the common candidate list using the common candidate list, and can be encoded/decoded using the same candidate generated.

[00478] Each region in which encoding/decoding is performed independently or in parallel may consist of a plurality of blocks, and prediction may be performed on each block. Here, prediction can be performed by generating corresponding lists of candidates in each area.

[00479] For example, when tiles A and tiles B constituting one picture are encoded/decoded in parallel, each region may be encoded/decoded using candidate lists different from each other. For example, an encoder or decoder may store motion information generated in a block belonging to tile A in a candidate list of tile A. Additionally, the encoder or decoder may store motion information generated in a block belonging to tile B in a candidate list of tile A. element B.

[00480] The candidate list newly generated by each region for independent or parallel encoding/decoding can be initialized or followed by accessing or copying the candidate list that is preconfigured. Once initialized, each region's candidate list can only be updated with information in each region. Accordingly, each region can be encoded/decoded independently of each other. This list of candidates is then defined as a sub-list of candidates.

[00481] FIG. 39 shows an example in which one or more tiles composed of one or more encoding blocks are encoded/decoded in parallel. The encoder or decoder may follow the reconstructed common candidate list for each tile as the initial value of each tile's candidate list. Sub-candidate lists that differ from each other can be generated for each tile based on the previously inherited candidate list.

[00482] Here, at least one of the information (flag, index, pointer, etc.) indicating whether to follow the reconstructed common candidate list as the initial value of the candidate list of each tile, the referenced picture information for initialization, tile information referenced for initialization, and information about the number of pieces of motion information referenced for initialization may be signaled by at least one of a sequence parameter set, a picture parameter set, and a slice header.

[00483] Referring to FIG. 39, the encoder or decoder may generate corresponding candidate sub-lists for tile 0 and tile 1 following the reconstructed overall candidate list. The first encoding block included in tile 0 may be encoded/decoded using the legacy candidate list. Here, the candidate list of tile 0 can be updated using the output motion information E. After the encoding/decoding of the first encoding block is completed, the motion information E of the first block can be added to the sub-candidate list.

[00484] Then, the second tile 0 encoding block may be encoded/decoded with reference to the updated candidate sub-list including the motion information E. The encoder or decoder may output the motion information F of the block and perform prediction for the block based on the motion information F Here, the candidate sub-list of tile 0 can be updated again using the derived motion information F.

[00485] FIG. 39 illustrates a case in which the size of the candidate list is 5, and shows an example in which one piece of previously stored motion information A is deleted, and motion information F is added to the candidate list according to the update of the motion information. Here, the motion information removed from the candidate sub-list may be motion information having the lowest or highest index, or motion information having the lowest prediction potential.

[00486] The encoder or decoder encodes/decodes the first encoding block using the inherited tile 1 candidate list and can update the tile 1 candidate list with motion information G obtained from the first encoding block that is encoded/decoded. Accordingly, the candidate sub-list updated after completion of encoding/decoding of the first encoding block further includes motion information G of the first block. Even when following the overall candidate list after the prediction for each tile is performed, the sub-candidate lists of tile 0 and tile 1 may include motion information different from each other.

[00487] Then, the second encoding block of tile 1 can be encoded/decoded with reference to the updated candidate sub-list including motion information G. Here, the candidate sub-list of tile 1 can be updated again using the outputted motion information G.

[00488] Since G motion information is already added to the tile 1 candidate sub-list by previously encoded/decoded motion information, adding redundant G motion information can be omitted. As described above, a sub-candidate list of each tile following the overall candidate list may be generated in a different form when each encoding block is encoded/decoded. Since cross-referencing between different tiles is not required, each tile can be encoded/decoded independently or in parallel.

[00489]

[00490] Next, a method for improving the candidate list prediction efficiency described above by configuring the UPU will be described.

[00491] Using the candidate list in accordance with the present invention, motion information generated in the previous block can be stored for accumulation and subsequent management. The motion information accumulated in the candidate list improves the encoding efficiency so as to address motion information of an area that cannot be accessed by the existing prediction method.

[00492] However, motion information generated at a position that is too distant has a problem that the prediction efficiency may be reduced in predicting the current block. In more detail, when the size of the candidate list is limited, a situation in which only motion information generated at a position too distant is stored in the candidate list may cause a decrease in encoding efficiency. Here, proper UPU configuration can solve or mitigate this problem.

[00493] FIG. 40 is a diagram illustrating a candidate list consisting of a single UPU and a candidate list consisting of multiple UPUs, in accordance with an embodiment of the present invention.

[00494] FIG. 40(a) shows an example of a candidate list structure consisting of one UPU. Fig. 40 shows an example in which the UPU is allocated based on the picture and the size of the candidate list is limited to eight. The encoding order is from left to right, and a large number of block divisions occur at the right end of the picture, and the motion information of the rightmost eight encoding blocks is stored in the candidate list. Here, the current block is present at the left end, so that the current block is too far from the encoding block referenced by the motion information stored in the candidate list, which reduces the encoding efficiency.

[00495] FIG. 40(b) illustrates an example of a UPU configuration to which an embodiment in accordance with the present invention is applied. The encoder or decoder can split the picture vertically so that the picture includes n UPUs. Here, since each UPU has a separate candidate list, the current block can be encoded/decoded with reference to the candidate list of UPU 0. According to the configuration of the UPU, even when the row of encoding blocks changes, the distance between the current block and the encoding block referenced by the motion information is may be closer than that in FIG. 40(a). Similarly, when a future block is encoded, the candidate list of UPU n is referenced, thereby preventing the case of accessing an encoding region that is too far away.

[00496] That is, the encoder or decoder may use different candidate lists according to the UPU that owns the location of the current block. Accordingly, the encoder or decoder can prevent deterioration in encoding efficiency. Here, the candidate list belonging to each UPU may reference a previously generated candidate list belonging to UPUs different from each other. For example, a candidate list of UPU n may be generated with reference to at least one of the UPUs UPU 0 to UPU n-1.

[00497]

[00498] FIG. 41 is a flowchart illustrating a decoding method according to an embodiment of the present invention.

[00499] With reference to FIG. 41, the decoding method according to the present invention may include deriving a first candidate list using the motion information of an adjacent block (S4110), deriving a second candidate list using the previously reconstructed motion information (S4120), deriving a third candidate list using the first candidate list and the second the candidate list (S4130); and generating a prediction block using the third candidate list (S4140).

[00500] Here, the previously recovered motion information may be the motion information of a decoded encoding block just before decoding the current block.

[00501] FIG. 42 is a flowchart illustrating an encoding method in accordance with an embodiment of the present invention.

[00502] With reference to FIG. 42, the encoding method according to the present invention may include determining motion information of the current block (S4210), outputting a first candidate list using motion information of an adjacent block (S4220), and outputting a second candidate list using the encoded motion information (S4230), and outputting a third list. candidates using the first candidate list and the second candidate list (S4240).

[00503] At least one or a combination of at least one of the above-described embodiments may be used to perform the encoding/decoding method shown in FIG. 41 and 42. For example, the first candidate list may refer to an existing candidate list, and the second candidate list may refer to a candidate list in accordance with the present invention. That is, the first candidate list may be at least one of the candidate lists used in the AMVP mode, the merge mode, and the IBC mode.

[00504] Here, the pre-coded motion information may be the motion information of an encoding block encoded immediately before encoding the current block.

[00505] Next, a method for encoding/decoding an image using a list of adaptive intra-picture prediction modes will be described.

[00506] In image compression, six or more bits of data are required to represent 32 or more in-picture prediction modes. Seven or more data bits are required to represent 64 or more intra-prediction modes. The image encoding/decoding method uses the most probable mode (MPM) to reduce the number of bits representing intra-prediction modes.

[00507] FIG. 43 and 44 are diagrams illustrating MPM candidates.

[00508] FIG. 43 is a diagram illustrating an example of a candidate that can be used as a current block and an MPM, wherein the gray shaded blocks corresponding to C1 to C17 are areas in which a neighboring block exists and can be used as an MPM. In image compression technology, the encoder configures the MPM candidate and the MPM list by deriving the intra-prediction mode from neighboring blocks of the current block, and when the same prediction mode as the intra-prediction mode defined in the current block is present in the MPM list, it can be reduced number of intra-prediction mode representation bits by passing the MPM candidate index instead of the current block's intra-prediction mode.

[00509] FIG. 43 shows an example of a candidate that can be used in the MPM, all or part of the candidates C1 to C17 can be used in the MPM according to the image compression method and according to the neighbor block prediction mode, at least one of the DC prediction mode, planar can be added to the MPM list prediction mode and directed intra-prediction mode. In addition, a directional mode obtained by adding or subtracting an offset to/from the intra-prediction directional mode(s) of a neighboring block may be added to the MPM list.

[00510] Additionally, as shown in the example of FIG. 44, the MPM list can be configured by deriving the intra-prediction mode from a block of an adjacent region instead of an adjacent region of the current block.

[00511] FIG. 45 is a diagram illustrating an example of intra-prediction mode selection, and FIG. 46 is a diagram illustrating a case where the intra-prediction mode of the current block is not included in the MPM list.

[00512] FIG. 45 is a diagram showing an example of the intra-prediction mode of the current block and the neighboring block in the image encoding/decoding process, wherein the gray shaded blocks represent regions included in the neighboring block, and the numbers for each region indicate the intra-prediction mode value of the neighboring region , including the corresponding area. When neighboring blocks of the current block, as shown in FIG. 45 do not include the same prediction mode as the current block, and when the intra-prediction mode of the current block is not included in the MPM list configured as shown in FIG. 46, the MPM for the current block is not selected, but the prediction mode of the current block can be directly signaled and transmitted.

[00513] As an example, when MPM is not selected as a method for compressing and transmitting the prediction mode of the current block, there is a method for transmitting it by classifying the prediction mode of the current block into a selected mode and a non-selected mode (remaining mode). For the remaining intra-prediction mode, excluding the intra-prediction mode numbers occurring for the MPM candidate in all possible intra-prediction modes, the selected mode compresses and transmits the selected mode flag and the intra-prediction mode number 1/N when the intra-prediction mode is a multiple of N (N is an integer greater than or equal to 2), and compresses and transmits the unselected mode along with the selected mode flag if it is not a multiple of N. Meanwhile, in the case of using the unselected mode, the encoder can perform transmission by encoding the intra-mode prediction (encoding the remaining intra-mode) that is not included in the MPM, using a method such as truncated binary encoding.

[00514] In addition, when the MPM is not selected, log N representation bits are required when representing N prediction modes (N is an integer equal to 2 or more) to represent the intra-prediction mode of the current block. In addition, there is a problem that more bits are generated than the MPM index even if the selected mode method is used.

[00515] In accordance with the present invention, it is possible to increase the MPM selectivity using a list of intra-prediction modes. According to some embodiments of the present invention, a method is provided for signaling an index of an adaptively configured intra-prediction mode list when an MPM is not selected, and a method for applying the intra-prediction mode list to the selected mode and the unselected mode to shorten the prediction mode representation bit. Since the list of intra-prediction modes in accordance with the present invention can always be accessed by the image encoder and decoder without additional signaling, the list of intra-prediction modes can be used as reference information in encoding/decoding processes other than the above-described MPM, selected mode and unselected mode.

[00516]

[00517] Next, a method for configuring a list of intra-prediction modes in accordance with the present invention will be described.

[00518] After performing intra-prediction by applying various intra-prediction modes, the encoder determines the prediction mode having the highest compression ratio among them as the optimal prediction mode, and the image decoder performs prediction and reconstruction using the intra-prediction mode determined by the encoder. Since the image encoder and decoder perform prediction and reconstruction using the same prediction mode, the image decoder must also be able to output the intra-prediction mode list output from the encoder. Since the decoder performs the same prediction through the intra-prediction mode determined by the encoder, when the rules for configuring the intra-prediction mode list of the encoder and the decoder are the same, the encoder and the decoder can configure the intra-prediction mode list in the same way without additional signaling. A list of intra-prediction modes configured in the encoder and decoder can be used when deriving the prediction mode.

[00519] FIG. 47 is a diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[00520] FIG. 47 shows an example of a process for configuring a list of intra-prediction modes in an encoder and decoder. According to FIG. 47, the list of intra-prediction modes can be configured based on the frequency of occurrence of the intra-prediction mode. In fig. 47, each of the squares constituting 16 blocks represents a unit encoding block outputting each prediction mode, the shadings in gray represent the current block, and the values in the shadings mean the average values updated in the count list counting the frequency of occurrence of each intra-prediction mode (intra-prediction mode). mode).

[00521] The index in the count list indicates the number of intra-prediction modes. Here, the number of intra-prediction modes may mean a value representing the intra-prediction mode. That is, in FIG. 47, the number in each prediction unit represents a value indicating the intra-prediction mode, and the dotted arrow represents the time flow. That is, intra prediction may be performed on the current block in the order (a) to (i) in FIG. 47.

[00522] Since the value of the intra-prediction mode defined in FIG. 47(a) is equal to five, the score list intra-prediction mode value 5 is increased by one. Then, since the value of the intra-prediction mode defined in FIG. 47(b) is equal to four, the score list intra-prediction mode value 4 is updated. Likewise, the value of the intra-prediction mode in the count list can be increased according to the intra-prediction mode defined in FIG. 47(c)-(i), respectively. The intra-prediction mode list used to signal the prediction mode in intra-prediction may be separately configured with reference to a configured count list or may be implemented to directly reference the count list.

[00523] FIG. 47 illustrates an example of an intra-prediction mode list output process, and an encoder or decoder may configure the intra-prediction mode list by calculating costs in a method other than counting. In addition, the method for listing the intra-prediction modes may be implemented differently. For example, costs can be determined using the residual signal for each intra-prediction mode. As another example, a list of intra-prediction modes may be configured by combining the frequency of occurrence and the cost result of the intra-prediction mode.

[00524] FIG. 48 is another diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[00525] FIG. 48 illustrates an example of configuring an intra-prediction mode list according to a specific intra-prediction mode. In the example in FIG. 48, since adjacent blocks adjacent to the top and left of each block and the corresponding blocks are different from each other in the prediction mode, MPM cannot be used. Accordingly, according to the method in the prior art, the encoder must signal all prediction modes, and when 35 modes are used, 5 bits must be allocated to each prediction mode to represent 32 modes that are not included in the MPM list. Therefore, to encode/decode the region in FIG. 48, for every 5 bits in 16 blocks, a total of 80 bits must be signaled.

[00526] Meanwhile, in accordance with the present invention, an encoder or decoder can configure a list of intra-prediction modes of a predetermined length. When the intra-prediction mode for the current block is present in the intra-prediction mode list, the encoder or decoder may reduce the number of bits required by index signaling in the intra-prediction mode list for the intra-prediction mode.

[00527] FIG. 48 shows an example of configuring an intra-prediction mode list having a length of 4 based on a count, which is the occurrence frequency of the prediction mode. After updating the score for each prediction mode, the encoder or decoder assigns an intra-prediction mode list index from value 0 in order from the highest score (in order of intra-prediction mode values 4, 3, and 5), thereby configuring the intra-prediction mode list. In addition, the encoder or decoder may generate a list of intra-prediction modes by adding or subtracting a predetermined weight to/from the count(s) for some prediction modes.

[00528] In the example of FIG. 48, when only the index for an intra-prediction mode list is signaled, since the index of a list of length 4 is signaled, the encoder or decoder can signal the intra-prediction mode using 2 bits per block, and signal the intra-prediction mode using 32 bits for 16 blocks. That is, by using the intra-prediction mode list according to the present invention, it is possible to improve the image encoding efficiency by greatly reducing the prediction mode representation bits.

[00529] FIG. 49 is a diagram illustrating a method for configuring a score-based intra-prediction mode list in accordance with an embodiment of the present invention.

[00530] When the count values are the same for all intra-prediction modes, the encoder or decoder sequentially configures the intra-prediction mode until the list of intra-prediction modes is full, or uses a prediction mode similar to the neighboring block's prediction mode to configure intra-prediction mode.

[00531] FIG. 49 shows an example of a method for configuring an intra-prediction mode list when the same score appears in a score-based intra-prediction mode list. With reference to FIG. 49, the score for values 2, 3, 4, 5 and 6 of the intra-prediction mode may have the highest value of the three in an identical manner. In this case, as shown on the left in FIG. 49, the encoder or decoder scans sequentially from 0 so that prediction modes generated with the same count can be added to the intra-prediction list until all intra-prediction mode lists are filled.

[00532] Alternatively, as shown on the right in FIG. 49, the encoder or decoder preferably adds the prediction modes of the left and top adjacent blocks of the current block to the prediction mode having the same count value, thereby configuring the intra-prediction mode list. Even if the score does not have the highest value, the encoder or decoder can apply priority to the prediction modes of the left and top adjacent blocks with respect to the prediction mode having the same score.

[00533] FIG. 50 is a diagram illustrating a method for configuring a cost-based intra-prediction mode list in accordance with an embodiment of the present invention.

[00534] As another example, even if the count is the same, the encoder or decoder may configure the intra-prediction mode using a predefined condition. For example, the predetermined condition may be the frequency of occurrence of the prediction mode of adjacent or neighboring blocks of the current block. The list of intra-prediction modes can be configured by preferentially using the prediction mode having the most optimal cost by calculating the cost for each prediction mode.

[00535] The cost measure may be calculated using the frequency of occurrence of each prediction mode, the residual signal generated for each prediction mode, and the like. for the intra-prediction mode selected as the optimal prediction mode. Additionally, the cost may be calculated by applying a weight to the intra-prediction mode defined in a neighboring block adjacent to the current block, or applying a weight to a direction similar to the intra-prediction mode occurring in a neighboring block adjacent to the current block.

[00536] FIG. 50 shows an example of a method for configuring a cost-based intra-prediction mode list. The shaded part of Fig. 50 means the current block, and each number can mean intra-prediction mode. In fig. 50, the intra-prediction mode occurring in each block can be stored in the history mode list. With reference to FIG. 50, the cost list may be configured using the absolute value of the difference between the prediction modes of the history mode list as a cost, based on the average of the prediction modes of blocks adjacent to the top and left of the current block as a reference value. Here, the encoder or decoder may configure a list of intra-prediction modes by preferentially adding a prediction mode having low cost

[00537] When configuring an intra-prediction mode list, when a newly added intra-prediction mode exists in an existing intra-prediction mode list, the index of the corresponding intra-prediction mode can be set to low. The index value can be used as an identifier to select one of the intra-prediction modes included in the list of intra-prediction modes. Here, the encoder or decoder can improve signaling efficiency since fewer bits signaled to the decoder are assigned as the index value decreases.

[00538] FIG. 51 and 52 are diagrams illustrating a method for changing an intra-prediction mode list index in accordance with some embodiments of the present invention.

[00539] Because the intra-prediction mode of the current block is more likely to be similar to the intra-prediction mode of the next block than the intra-prediction mode of previously encoded blocks, when adding the current block's intra-prediction mode to the list of intra-prediction modes, the lowest one may be assigned index value “0”.

[00540] FIG. 51 shows an example of a method of applying the lowest index to an intra-prediction mode when the added intra-prediction mode exists in an existing list of intra-prediction modes. According to FIG. 51, when intra-prediction mode values 0, 22, 3, 11 and 7 are present in the intra-prediction mode list, and the added intra-prediction mode is 3, the encoder or decoder may change the value of intra-prediction mode index 3 which is already present in the list of intra-prediction modes, to zero.

[00541] FIG. 52 shows an example of a method for reducing an index for an intra-prediction mode when the added intra-prediction mode exists in an existing list of intra-prediction modes. When there are intra prediction mode values 0, 22, 3, 11 and 7 in the intra prediction mode list and the added intra prediction mode is three, the encoder or decoder can change the index of intra prediction mode 3 already present in the intra prediction mode list. predictions, from third to second. Alternatively, when configuring a list of intra-prediction modes, when a newly added intra-prediction mode exists in an existing list of intra-prediction modes, the encoder or decoder may assign a predefined index to the corresponding intra-prediction mode.

[00542] FIG. 53 and 54 are diagrams illustrating an area configuring an intra-prediction mode list in accordance with some embodiments of the present invention.

[00543] A list of intra-prediction modes may be configured using a specific encoding/decoding unit, such as block, CTU, brick, tile, slice, sub-picture, picture, frame, and group of pictures (GOP), composed of one or more CUs . Because adjacent blocks have a higher probability of generating the same prediction mode, the encoder or decoder may configure a list of intra-prediction modes in units such as CTU or slice. When there are not enough candidates for intra-prediction, a list of intra-prediction modes is configured in picture units, thereby outputting more accurate predictions. In addition, since a picture has high similarity between pictures, the encoder or decoder can improve coding efficiency by configuring a list of intra-prediction modes for each GOP. In addition, since adjacent blocks have a relatively high probability of generating the same prediction mode, the encoder or decoder can generate a list of intra-prediction modes for each predefined region in the picture.

[00544] Here, the predefined area may be limited to an adjacent area of the current block, as shown in FIG. 53, or can be formed by designating an area by performing N-partitioning on the screen (N is an integer equal to 1 or more), as shown in FIG. 54. A list of intra-prediction modes in accordance with the present invention is generated and used not only for a block, CTU, brick, tile, slice, sub-picture, picture, frame, group of picture unit (GOP) unit consisting of one or more CUs, but also the various units that can be used in the field of video compression, including the still image compression method. In addition, the list of intra-prediction modes may be configured for all prediction modes or may be configured as a separate list only for a prediction mode having a high probability of selection.

[00545] FIG. 55 is another diagram illustrating a method for configuring an intra-prediction mode list in accordance with an embodiment of the present invention.

[00546] FIG. 55 shows another example of a method in which an image encoder or decoder configures a list of intra-prediction modes. In the example in FIG. 55, when the intra-prediction mode list is full, the intra-prediction mode list may be configured with a first-in, first-out (FIFO) rule that replaces the first added component with a new value. When the list of intra-prediction modes is full, the replacement order can be changed taking into account the intra-prediction direction of the adjacent block.

[00547] In the following description, a list of intra-prediction modes in which the replacement order follows the first-in-first-out (FIFO) rule will be described. In fig. 55, each of the squares constituting 16 blocks may indicate a unit block for outputting a prediction mode, and shading in gray may indicate the current block. The numbers in each prediction unit indicate the intra-prediction mode, and the dotted arrows indicate the intra-prediction order. That is, intra-prediction may be performed in the order of (a) to (f) in FIG. 55.

[00548] Since the intra prediction mode value defined in FIG. 55(a) is five, mode number five can be added to the list of intra-prediction modes. Then, since the value of the intra-prediction mode defined in FIG. 55(b) is equal to four, mode number 4 can be successively added to the list of intra-prediction modes. Likewise, the intra-prediction modes respectively defined in FIG. 55(c)-55(f) can be successively added to the list of intra-prediction modes. When the intra-prediction mode added in the process of adding each intra-prediction mode to the intra-prediction mode list exists in the intra-prediction mode list, the overlapping intra-prediction mode cannot be added to the intra-prediction mode list.

[00549] Additionally, as shown in FIG. 55(f), when the intra-prediction mode lists are full, each component may be successively replaced by a new intra-prediction mode in the order of first addition among the components of the intra-prediction mode list. In fig. 55, the size of the intra-prediction mode list is four, but the size of the intra-prediction mode list can be set in different ways. As another example, when the list of intra-prediction modes is full, the added intra-prediction mode may be added to a predetermined position in the list of intra-prediction modes.

[00550] The intra-prediction mode list index may be used as an identifier indicating one of the intra-prediction modes included in the intra-prediction mode list. Here, the encoder or decoder can improve signaling efficiency by assigning fewer bits to be signaled to the decoder when the index value is smaller.

[00551] When the intra-prediction mode of the current block is more likely to be similar to the intra-prediction mode of the next block than the intra-prediction mode of previously encoded blocks, when the intra-prediction mode is added to the list of intra-prediction modes, the encoder or decoder may assign the lowest index to the added intra-prediction mode. For example, in FIG. 55(b), the encoder or decoder changes the intra-prediction mode value of five, which is the value of the existing index No. 0, to the index value of No. 1, and the intra-prediction mode value of 4 in the current encoding block can be assigned to the index value No. 0.

[00552] FIG. 56-58 are diagrams illustrating a method for allocating an intra-prediction mode index in accordance with some embodiments of the present invention.

[00553] FIG. 56 shows an example of a method for assigning the lowest index value to an added intra-prediction mode. The lowest index value of 0 may be assigned to the last added intra-prediction mode, and lower index values may be assigned to each intra-prediction mode in the order of later added intra-prediction modes for the remaining intra-prediction mode numbers.

[00554] As another example, an encoder or decoder may sequentially sum the output intra-prediction mode, reorder the index values, and assign a value of “0” starting with the last added intra-prediction mode. For example, in FIG. 55(f), the list of intra-prediction modes can be configured such that the values of the intra-prediction modes are in the order of 4, 3, 6 and 8, and here the index values are in the order of 3, 2, 1 and 0, in which the mode Intra-prediction mode 4 may be assigned an index value of 3, intra-prediction mode 3 to an index value of 2, intra-prediction mode 6 to an index value of 1, and intra-prediction mode 8 to an index value of 0, respectively.

[00555] FIG. 57 shows an example of a method for assigning the lowest index value to a newly added mode. The lowest index value of zero may be assigned to the most recently added intra-prediction mode, and lower index values may be assigned in the order of the most recently added intra-prediction modes for the remaining intra-prediction modes.

[00556] When the added intra-prediction mode already exists in the list of intra-prediction modes, the intra-prediction mode may be assigned “zero”, which is the lowest index value. Fig. 58 illustrates an example of a method for assigning the lowest index value in the intra-prediction mode list to an intra-prediction mode when the intra-prediction mode already exists in the intra-prediction mode list. With reference to FIG. 58(d), in the process of adding the intra-prediction mode value 4, since the intra-prediction mode 4 already exists in the intra-prediction mode, the encoder or decoder may assign zero, which is the lowest index value for the intra-prediction mode value 4.

[00557]

[00558] Next, a method for configuring a plurality of intra-prediction mode lists according to an embodiment of the present invention will be described.

[00559] There may be multiple lists of intra-prediction modes in a certain area. In particular, when multiple intra-prediction mode lists are configured for each direction or each prediction mode region, it is advantageous to take into account the directionality of the prediction mode, the pixel characteristics of a certain region, etc., rather than configuring one intra-prediction mode list for all prediction modes and all areas.

[00560] FIG. 59 is a diagram illustrating a method for configuring a plurality of intra-prediction mode lists in accordance with an embodiment of the present invention.

[00561] FIG. 59 shows an example of a method for configuring one intra-prediction mode list and a method for configuring multiple intra-prediction mode lists for each region. Fig. 59(a) illustrates an example of configuring one intra-prediction mode list having a length of 2. FIG. 59(b) illustrates an example of configuring an intra-prediction mode list having a length of 2 for a left region (first region) and a right region (second region) relative to the block center.

[00562] In FIG. 59(a), since one list of intra-prediction modes is configured without distinguishing between left and right regions, the characteristic of generating similar intra-prediction modes for each region may not be taken into account. Meanwhile, in FIG. 59(b), the list of intra-prediction modes can be individually configured based on the position and orientation for each region relative to the center. When configuring multiple intra-prediction lists for each region, the intra-prediction mode generation characteristics of each region can be taken into account, thereby improving coding efficiency.

[00563] In the example of FIG. 59, the block is divided into left and right areas, but the embodiment of FIG. 59 is only an example in which a predefined area can be divided into N arbitrary areas (N is an integer equal to 1 or more) to generate N intra-prediction lists. Here, the predefined area may be limited to the peripheral area of the current block or may mean an area obtained by performing N-partitioning on the screen (N is an integer equal to 1 or more) into slices, pictures, and the like. instead of blocks. In addition, the encoder or decoder may use the intra-prediction mode of a partial region to configure the intra-prediction mode of another partial region.

[00564] FIG. 60 is a diagram for describing a method for configuring a plurality of intra-prediction mode lists in consideration of prediction mode directionality in accordance with an embodiment of the present invention.

[00565] FIG. 60 shows an example of a method for configuring one intra-prediction mode list and a method for configuring multiple intra-prediction lists for each direction. Fig. 60(a) illustrates an example of configuring one intra-prediction mode list having length two, and FIG. 60(b) illustrates an example of a method for configuring an intra-prediction mode list having a length of two for a horizontal direction and a vertical direction based on the directionality of the prediction mode.

[00566] Because prediction modes having a similar direction in intra-prediction have characteristics that appear in a contiguous manner, the encoder or decoder generates a list of intra-prediction modes for each direction and then searches for the prediction mode of the current block in the list of intra-prediction similar the direction of the neighboring block, taking into account the direction of the prediction mode of the neighboring block of the current block. By selecting a list of intra-prediction modes based on directionality, the ability to find the same intra-prediction mode can be increased. Accordingly, the configuration of the prediction mode list for each direction as shown in FIG. 60, can improve coding efficiency.

[00567] In the example of FIG. 60, the prediction direction is classified into horizontal and vertical directions, but the embodiment of FIG. 60 is only an example in which the direction can be classified in various ways, so that N or more (N is an integer of 1 or more) intra-prediction lists can be configured.

[00568] FIG. 61 illustrates a method for discriminating direction when configuring N or more intra-prediction lists. Fig. 61(a) illustrates an example of a method for distinguishing two directions, and FIG. 61(b) illustrates an example of a method for distinguishing four directions.

[00569] FIG. 62 is a diagram illustrating an encoder and decoder structure that complements a list of intra-prediction modes in accordance with an embodiment of the present invention.

[00570] The device diagram of FIG. 62 may have an intra-prediction mode list generator added to the device diagrams of FIG. 1 and 2. The intra-prediction mode list generator may configure the intra-prediction mode list to reduce the number of intra-prediction mode signaling bits used for intra-prediction. In addition, the encoder or decoder may employ a method of using an intra-prediction mode list component as an MPM candidate without signaling the intra-prediction mode list index.

[00571] FIG. 63 is a diagram illustrating a method for expanding a list of intra-prediction modes in accordance with one embodiment of the present invention.

[00572] FIG. 63 shows an example of a method for configuring a list of intra-prediction modes in an encoder and decoder. Fig. 63(a) shows a flowchart in which the encoder configures a list of intra-prediction modes. When the intra-prediction mode of the current block is determined in the encoder, the encoder may add the specific intra-prediction mode to the list of intra-prediction modes. Here, when the list of intra-prediction modes is full, the intra-prediction mode can be added according to the first-in-first-out (FIFO) rule, which replaces the first added component with a new value.

[00573] When the list of intra-prediction modes is full, the replacement order can be partially changed, as shown in FIG. 64 taking into account the intra-prediction mode of the neighboring block. With reference to FIG. 64(h), in the process of adding intra-prediction mode value 4, intra-prediction mode value 6 of index No. 0 must be replaced by 4 in FIFO order. However, since the intra-prediction mode values of the adjacent left and top blocks of the current block are 5 and 7, respectively, the encoder or decoder replaces the intra-prediction mode 34 having the largest difference from the average value 6 of intra-prediction mode values 5 and 7 in the mode list intra-prediction mode to intra-prediction mode value 4, thereby configuring an intra-prediction mode list reflecting the prediction mode trend of the peripheral region of the current block. Fig. 64 illustrates an example of a substitution order changing method in which the criteria for changing the substitution order may be determined not only by the average value of the intra-prediction modes of neighboring blocks, but also by the average value of neighboring blocks of non-adjacent current blocks, as shown in FIG. 64. In addition, the intra-neighbor block prediction mode can be used as a support for changing the replacement order. In addition, at least one of the statistical values, such as a weighted sum, a weighted average, a median, a minimum value, a maximum value, etc., in addition to the average value, can be used to determine criteria for changing the substitution order.

[00574] FIG. 63(b) shows a flowchart in which the decoder configures a list of intra-prediction modes. The decoder may decode the prediction mode signaled from the encoder and add it as a component of the intra-prediction mode list. When the list of intra-prediction modes is full, an intra-prediction mode can be added according to a first-in, first-out (FIFO) rule, which replaces the first component added with a new value, as in an encoder.

[00575] When the list of intra-prediction modes is full, the substitution order may be partially changed taking into account the intra-prediction mode of an adjacent block.

[00576] FIG. 65 is a diagram illustrating a method of applying an intra-prediction mode list to an MPM in accordance with an embodiment of the present invention.

[00577] FIG. 65 shows an example of a method for accessing the intra-prediction mode list when configuring the MPM list in an encoder and decoder. Fig. 65(a) shows an example of a method for deducing all or some MPM components from a list of intra-prediction modes. When using a 3 MPM scheme to configure an MPM list using three MPMs, the encoder or decoder can derive the intra-prediction mode from the top and left adjacent blocks.

[00578] Here, when the intra-prediction modes of two adjacent blocks are the same, the MPM list component is configured through the intra-prediction mode of the adjacent block, and one or two components of the remaining two components can be determined using the components in the intra-prediction mode list. In addition, even when the intra-prediction modes of two adjacent blocks are different, another intra-prediction mode can be determined using the components of the intra-prediction mode list. Likewise, even when an MPM list of six MPMs is configured, partial MPM list candidates can be determined through the intra-prediction mode candidate list.

[00579] Additionally, as shown in FIG. 65(b), when the intra-prediction mode cannot be derived from an adjacent block of the current block, the MPM list can be configured using the components of the intra-prediction mode list. Although the above description describes that some MPM configuration conditions are applicable to the embodiment of the present invention, this method does not mean that the present invention is limited to only 3 MPM and 6 MPM. To configure different types of MPM lists, an intra-prediction mode list can be used in accordance with the present invention.

[00580] FIG. 66 is a diagram illustrating a method of applying a list of intra-prediction modes to a non-MPM mode in accordance with an embodiment of the present invention.

[00581] FIG. 66 shows an example of a method for accessing an intra-prediction list in an encoder in the case of a non-MPM. The encoder can check whether the current intra-prediction mode is included in the MPM list. When the current intra-prediction mode is included in the MPM list, the encoder encodes the MPM flag and the MPM index and transmits them to the decoder, and the encoding process for the intra-prediction mode can be completed.

[00582] When the encoder is operating in non-MPM mode, the MPM flag may be encoded to 0 and signaled to the decoder. When the current intra-prediction mode is not in the MPM list, the encoder may determine whether the current intra-prediction mode exists in the list of intra-prediction modes, and may signal an identifier (or flag, index, etc.) indicating whether to access to the list of intra-prediction modes.

[00583] When the current intra-prediction mode is in the intra-prediction mode list, the encoder may encode an identifier indicating whether to access the intra-prediction mode list to a first value (for example, “1”), and encode the index of the mode list intra-prediction that corresponds to the current intra-prediction mode to signal it to the decoder.

[00584] When the current intra-prediction mode is not in the intra-prediction mode list, the encoder may encode an identifier indicating whether to access the intra-prediction mode list into a second value (eg, “0”) to signal it to the decoder , and encode the current intra-prediction mode to signal it to the decoder.

[00585] In FIG. 66, since it is determined whether to select the intra-prediction mode list for a mode not selected in the MPM when operating in the non-MPM mode, the situation in which the intra-prediction mode present in the MPM list is the prediction mode of the current block is not arises. Therefore, in the case of a non-MPM mode according to the present invention, the list of intra-prediction modes can be configured using a mode excluding MPM components.

[00586] In a non-MPM encoding mode, mode encoding may be performed for the remaining modes except the MPM candidate. For example, in the case where 67 modes are used in the intra-prediction mode and 3 MPM is used, when the encoder or decoder operates in the non-MPM mode, encoding can be performed for the 64 modes excluding the three modes constituting the MPM from the 67 intra-prediction modes. predictions. According to the present invention, when an encoder or decoder has N prediction modes, using n1 (n1 is a natural number) MPM candidates and using a prediction mode derived from n2 (n2 is a natural number) intra-prediction mode lists, the encoder can perform CABAC encoding in N-n1-n2 modes.

[00587] In non-MPM mode, the encoder may use a list of intra-prediction modes even in the selected mode's mode. The selected mode is a method used in non-MPM mode, and an encoding method that distinguishes n (n is a natural number) selected modes and N-n unselected modes, in order to allocate fewer bits to the selected mode, for N modes (N is a natural number), which may appear in non-MPM mode,

[00588] When using an intra-prediction mode list, an intra-prediction mode list component may be added to the method of the selected mode. Here, since the selection frequency of the selected mode can be increased without additionally transmitting the intra-prediction mode list selection flag, the encoding efficiency can be increased. Alternatively, the encoder or decoder may configure the intra-prediction mode list in modes that are not multiples of a such that the intra-prediction mode list component does not overlap with the selected mode, thereby minimizing the occurrence of remaining modes.

[00589] FIG. 67 is a diagram illustrating a method of applying an intra-prediction mode list to a non-MPM mode by a decoder according to an embodiment of the present invention.

[00590] FIG. 67 shows an example of a method for accessing an intra-prediction list by a decoder in the case of a non-MPM mode. The decoder may receive and decode the MPM flag to identify whether to apply the MPM sent from the encoder. When the MPM flag is 1, the MPM index can be decoded to restore the intra-prediction mode from the MPM list and complete the intra-prediction mode restoration process.

[00591] In the case of a non-MPM mode in which the MPM flag is 0, in accordance with the present invention, the decoder can decode the intra-prediction mode list access flag. When the intra-prediction mode list access flag is 1, the decoder can decode the intra-prediction mode list index, restore the intra-prediction mode through the intra-prediction mode list index, and complete the intra-prediction mode restoration process.

[00592] When the intra-prediction mode list access flag is 0, the current intra-prediction mode can be decoded via CABAC. Here, the decoder operations can be replaced with each other in cases where the MPM flag is equal to 1 and 0. In addition, the decoder operations can be replaced with each other in cases where the intra-prediction mode list access flag is equal to 1 and 0.

[00593] In FIG. 67, since the intra-prediction mode list is selected for a mode not selected in the MPM, when the non-MPM mode is enabled, there is no case where the intra-prediction mode present in the MPM list is the same as the prediction mode of the current block. Therefore, in the case of a non-MPM mode according to the present invention, a list of intra-prediction modes can be configured using a mode excluding MPM components.

[00594] Mode decoding for non-MPMs must be performed in the same manner as for an encoder. Therefore, the method of configuring the intra-prediction mode list and the selected decoder mode that is associated with the selected mode and the non-selected mode can be performed in the same manner as in the above-described encoder. The list of intra-prediction modes can also be accessed and used in intra-prediction modes other than MPM, selected mode, and unselected mode.

[00595] FIG. 68 is a diagram illustrating an image encoding/decoding method using an intra-prediction mode list according to an embodiment of the present invention.

[00596] To configure a list of intra-prediction modes, the encoder or decoder may calculate costs for each prediction mode. Fig. 68(a) illustrates a flowchart of a method for configuring an intra-prediction mode list by an encoder, and FIG. 68(b) illustrates a flowchart of a method for configuring an intra-prediction mode list by a decoder. As shown in FIG. 68(a), when an intra-prediction mode is determined, the encoder may update the costs for the corresponding intra-prediction mode. Here, the cost may be a value derived using the occurrence frequency or residual signal of each intra-prediction mode. Costs can be configured in the form of a list or the like. and can be used as information for configuring a list of intra-prediction modes.

[00597] According to the present invention, when the intra-prediction mode list is configured based on the cost of each prediction mode, and the same intra-prediction mode as the current block's prediction mode is in the intra-prediction mode list, the encoder may not signal directly the intra-prediction mode, and signal the index of the list of intra-prediction modes. The decoder configures the intra-prediction mode list in the same manner as the encoder based on the information signaled from the encoder, and uses the intra-prediction mode list as reference information for reconstructing the prediction mode, as shown in FIG. 68(b). The list of intra-prediction modes can be used as reference information in various prediction methods associated with intra-prediction mode signaling.

[00598] FIG. 69 is a view illustrating a method for configuring an intra-prediction mode list according to an embodiment of the present invention.

[00599] In the method for configuring a list of intra-prediction modes based on the costs of each intra-prediction mode, an example of the cost calculation method is illustrated in FIG. 69. Costs may be calculated using at least one of a frequency-of-occurrence count or a residual signal after prediction. Fig. 69(a) illustrates a method for configuring a list of intra-prediction modes based on a count that is the occurrence frequency of each prediction mode, and FIG. 69(b) illustrates a method for calculating both a count and a residual signal and configuring a list of intra-prediction modes based on the average of the residual signals for each prediction mode.

[00600] Since a prediction mode that frequently occurs in intra-prediction is more likely to be reused, the list of intra-prediction modes can be configured by giving priority to a prediction mode having a high score, according to FIG. 69(a).

[00601] Meanwhile, since a less accurate residual signal is generated during intra-prediction execution, more accurate prediction is performed, the list of intra-prediction modes can be configured by giving priority to a prediction mode having a low average value of the residual signals, according to FIG. 69(b).

[00602] The cost used in the method for configuring the list of intra-prediction modes can be calculated according to all the criteria that can configure the list of intra-prediction modes, such as the method of checking the frequency of occurrence of each prediction mode for neighboring regions of the current block.

[00603] FIG. 70 and 71 are diagrams illustrating an encoding/decoding method in accordance with some embodiments of the present invention.

[00604] The list of intra-prediction modes may not be configured separately, as in the example above, but may also be configured in a manner that directly refers to the costs calculated for each prediction mode. In this case, the flowchart is as shown in FIG. 70. According to another embodiment of the present invention, as shown in FIG. 71, a list of intra-prediction modes can be configured without cost measurement. In this case, the list of intra-prediction modes may be adaptively configured using the intra-prediction mode of a neighboring block adjacent to the current block.

[00605] FIG. 72 is a diagram illustrating a method for initializing an intra-prediction list in accordance with an embodiment of the present invention.

[00606] The list of intra-prediction modes cannot be generated until the intra-picture prediction has been performed. Accordingly, the encoder or decoder may assign an initial value to the list of intra-prediction modes. Fig. 72 shows an example of an initialized intra-prediction mode list. An encoder or decoder may preferentially assign a low index to a mode that is used in most cases statistically for a list of intra-prediction modes.

[00607] In FIG. 72, PLANAR_IDX means a planar mode number for intra-prediction, DC_IDX means a DC mode number for DC-prediction, and VER_IDX is a mode number for directional intra-prediction in the vertical direction, and HOR_IDX means a mode number for directional intra-prediction in the horizontal direction. Statistically, when PLANAR_IDX, DC_IDX, VER_IDX and HOR_IDX occur frequently, in the example of FIG. 72, an encoder or decoder may assign indices from 0 to 3 to PLANAR_IDX, DC_IDX, VER_IDX, and HOR_IDX. The encoder or decoder may then sequentially add modes, by adding and subtracting the offset values to/from the directional mode(s) VER_IDX and HOR_IDX, to the list of intra-prediction modes. In the example in FIG. 72, offset 1 and offset 2 correspond to offset values, and as the index in the list increases, the offset value can increase.

[00608] That is, in FIG. 72, offset 2 > offset 1. Thus, an encoder or decoder may initialize a list of intra-prediction modes using an intra-prediction mode that statistically has a high frequency of occurrence. A directional prediction mode having little difference from an intra-prediction mode having a high occurrence frequency can be set as an initial value, and the ability to access the list of intra-prediction modes can be increased.

[00609] Additionally, as an example, offset 2 may be determined by adding or subtracting a predetermined value to/from offset value(s) 1. In the example of FIG. 72, VER_IDX and HOR_IDX are set as directional modes for adding or subtracting offset. However, this is a value arbitrarily defined for convenience of description, and this value may vary depending on the encoding/decoding situation. In addition, as a method for determining the directional mode set as the initialization value, a method of using the first directional intra prediction mode number specified as the prediction mode can be used.

[00610] Table 1 below is a table showing representation bits that may indicate an intra-prediction mode in intra-prediction.

[00611] [Table 1]

[00612]

Embodiment 1
(35 modes) Embodiment 2
(67 modes) MRM flag 1 bit 1 bit MRM modes 2 bits (3 modes) 2 bits (3 modes) Non-MRM modes 5 bits (32 modes) 6 bits (64 modes)

[00613] According to Table 1, 1 bit may be allocated for the MPM flag, and 2 bits may be allocated to represent the three MPM modes. As representation bits for a non-MPM mode, five bits can be used in a 32-mode allocation, and six bits can be used in a 64-mode allocation.

[00614] Table 2 below is a table showing an example of an intra-prediction mode representation bit in accordance with the present invention.

[00615] [Table 2]

[00616]

Embodiment 1
(35 modes) Embodiment 2
(67 modes) MRM flag 1 bit 1 bit MRM modes 2 bits (3 modes) 2 bits (3 modes) Non-MRM modes Intra-mode list flag 1 bit 1 bit Intra-mode list modes 2 bits (4 modes) 3 bits (8 modes) Non-intra-mode list modes 5 bits (28 modes) 6 bits (56 modes)

[00617] In Embodiment 1 using 35 modes, the length of the intra-prediction mode list can be set to four, and in Embodiment 2 using 67 modes, the length of the intra-prediction mode list can be set to eight.

[00618] For convenience of description below, the intra-prediction mode list is defined as the intra-mode list, the flag indicating whether the intra-prediction mode list is selected as the intra-mode list flag, and the intra-prediction mode list mode as the modes list of intra-modes.

[00619] In Embodiment 1, 1 bit may be allocated to the intra-mode list flag, and 2 bits may be allocated to represent the four modes included in the intra-prediction mode list. Five bits may be allocated as bits to represent the remaining intra-prediction modes not included in the list of intra-prediction modes. In Example 1 of Table 1, 5 bits are permanently allocated for the mode in which MPM is not selected. However, when the intra-prediction mode list is used, the intra-prediction mode can be encoded using 3 bits including the flag, thereby improving encoding efficiency.

[00620] Similarly, in Embodiment 2, 1 bit may be allocated for the intra-mode list flag, and 3 bits may be allocated to represent the 8 modes included in the intra-prediction mode list. Six bits may be allocated as bits to represent the remaining intra-prediction modes not included in the list of intra-prediction modes. In Example 2 of Table 1, 6 bits are permanently allocated for the mode in which MPM is not selected. However, when the intra-prediction mode list is used, 4 bits including the flag can be allocated to the intra-prediction mode, thereby improving coding efficiency.

[00621] Table 3 shows an example of intra-prediction mode representation bits in which the number of MPM prediction modes is reduced and the size of the intra-prediction mode list is increased compared with Table 2. For example, the size of the MPM list and the size of the intra-prediction mode list in an encoder and decoder, to which a list of intra-prediction modes is added, can be set differently.

[00622] [Table 3]

[00623]

Embodiment 1
(35 modes) Embodiment 2
(67 modes) MRM flag 1 bit 1 bit MRM modes 1 bit (2 modes) 1 bit (2 modes) Non-MRM modes Intra-mode list flag 1 1 Intra-mode list modes 3 bits (8 modes) 4 bits (16 modes) Non-intra-mode list modes 5 bits (25 modes) 6 bits (49 modes)

[00624] Table 4 below is a table showing intra-prediction mode representation bits that can be used when using a plurality of intra-prediction mode lists.

[00625] [Table 4]

[00626]

Embodiment 1
(35 modes) Embodiment 2
(67 modes) MRM flag 1 bit 1 bit MRM modes 2 bits (3 modes) 2 bits (3 modes) Non-MRM modes Intra-mode list flag 1 bit 1 bit Intra-mode list modes 2 bits (4 modes) 3 bits (8 modes) Non-intra-mode list modes 5 bits (32 modes) 6 bits (64 modes)

[00627] In the first embodiment, the intra-prediction mode list may have a total of four lists for each direction, each length is four, and the plurality of intra-prediction mode lists may encode a total of 16 modes. In this case, the encoder or decoder may select one list from a plurality of intra-prediction mode lists based on the direction of the MPM or adjacent block and set the intra-mode list flag to 1.

[00628] Similarly, in Embodiment 2, the intra-prediction mode list may have a total of four lists for each direction, each length being eight, and the intra-prediction mode list may indicate a total of 32 modes. Likewise, the encoder or decoder may select one list from a plurality of intra-prediction lists based on the MPM direction or adjacent block and set the intra-mode list flag to 1.

[00629] Table 5 shows the encoding method of the intra-prediction mode, including the selected mode and the non-selected mode.

[00630][Table 5]

[00631]

Intra-prediction modes MRM flag Selected flag Bean string MRM modes (6) 1 0 10 110 1110 11110 11111 Selected modes (16) 0 1 4-bit fixed length code Unselected modes (45) 0 0 Truncated binary code

[00632] Table 5 shows the encoding method of the intra-prediction mode, including the selected mode and the non-selected mode. In the encoding method shown in Table 5, selected modes can be implemented using a list of intra-prediction modes to improve selectivity. In addition, the representation bit may be shortened by applying a method using the list of intra-prediction modes to unselected modes, or applying only the unselected modes to an intra-prediction mode not included in the list of intra-prediction modes.

[00633] FIG. 73 is a diagram illustrating the syntax of an intra-prediction mode list according to an embodiment of the present invention.

[00634] FIG. 73 shows an example of a syntax configuration including intra-prediction mode list information. A decoder in accordance with the present invention may signal Intra_mode_list_flag whether the intra-prediction mode list is selected for modes not included in the MPM. When Intra_mode_list_flag is equal to 1, the decoder can recover the intra-prediction mode by signaling Intra_mode_list_index, which is the index in the intra-prediction mode list.

[00635]

[00636] The above embodiments may be performed in the same manner in an encoder and a decoder.

[00637] At least one or a combination of the above embodiments may be used for video encoding/decoding.

[00638] The sequence applied to the above embodiment may be different for the encoder and decoder, or the sequence applied to the above embodiment may be the same in the encoder and decoder.

[00639] The above embodiment may be performed for each luma signal and chrominance signal, or the above embodiment may be identically performed for luma and chrominance signals.

[00640] The block shape to which the above embodiments of the present invention are applied may be a square shape or a non-square shape.

[00641] The above embodiment of the present invention can be applied depending on the size of at least one of a coding block, a prediction block, a transform block, a block, a current block, a coding unit, a prediction unit, a transform unit, a unit, and a current unit. Here, the size may be defined as a minimum size or a maximum size or both to apply the above embodiments, or may be defined as a fixed size to which the above embodiment is applied. Moreover, in the above embodiments, the first embodiment may be applied to the first size, and the second embodiment may be applied to the second size. In other words, the above embodiments can be used in combination depending on the size. Moreover, the above embodiments may be applied when the size is equal to or greater than the minimum size and equal to or less than the maximum size. In other words, the above embodiments can be applied when the block size falls within a certain range.

[00642] For example, the embodiments described above may be applied when the current block size is 8×8 or larger. For example, the above embodiments may be applied when the current block size is 4×4 or larger. For example, the above-described embodiments may be applied when the current block size is 16×16 or larger. For example, the above embodiments may be applied when the current block size is equal to or greater than 16×16 and equal to or less than 64×64.

[00643] The above-mentioned embodiments of the present invention can be applied depending on the time level. To identify the time layer to which the above embodiments may be applied, a corresponding identifier may be signaled, and the above embodiments may be applied to the specified time layer identified by the corresponding identifier. Here, the identifier may be defined as the lowest level or the highest level, or both, to which the above embodiment may apply, or may be defined to indicate a specific level to which this embodiment applies. In addition, a fixed time level to which the embodiment applies can be defined.

[00644] For example, the embodiments described above may be applied when the time level of the current image is the lowest level. For example, the above embodiments may be applied when the temporal layer ID of the current image is 1. For example, the above embodiments may be applied when the temporal layer of the current image is the highest level.

[00645] The slice type or tile group type to which the above embodiments of the present invention are applied can be determined, and the above embodiments can be applied depending on the corresponding slice type or tile group type.

[00646] In the above-described embodiments, the methods are described on the basis of flowcharts with a sequence of steps or modules, but the present invention is not limited to the order of the steps, and rather, some steps may be performed simultaneously or in a different order with other steps. In addition, one skilled in the art will appreciate that the steps in the flowcharts are not mutually exclusive and that other steps may be added to the flowcharts or some of the steps may be deleted from the flowcharts operations without affecting the scope of the present invention.

[00647] Embodiments include various aspects of the examples. All possible combinations for the various aspects may not be described, but those skilled in the art will recognize the various combinations. Accordingly, the present invention may include all substitutions, modifications and changes within the scope of the claims.

[00648] Embodiments of the present invention may be implemented in the form of program instructions that are executed by various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include, individually or in combination, program instructions, data files, data structures, etc. Program instructions recorded on a computer-readable recording medium may be specifically designed and constructed for the present invention or well known to one skilled in the art of computer software technology. Examples of computer-readable recording media include magnetic recording media such as hard disks, floppy disks and magnetic tapes; optical storage media such as CD-ROM or DVD-ROM; magneto-optical media such as floptic disks; and hardware devices such as read-only memory (ROM), random access memory (RAM), flash memory, etc., which are specifically structured to store and implement a software instruction. Examples of program instructions include not only mechanical language code formatted by a compiler, but high-level language code that can be implemented by a computer using an interpreter. Hardware devices may be configured to operate one or more software modules or, alternatively, to perform processes in accordance with the present invention.

[00649] Although the present invention has been described in terms of specific objects, such as detailed elements, as well as limited embodiments and drawings, they are provided only to facilitate a more general understanding of the invention, and the present invention is not limited to the above embodiments. Those skilled in the art to which the present invention relates will appreciate that various modifications and changes may be made based on the above description.

[00650] Therefore, the spirit of the present invention should not be limited to the embodiments described above, and the entire scope of the appended claims and their equivalents will be within the scope and spirit of the invention.

Industrial applicability

[00651] The present invention can be used to encode or decode an image.

1. A method for decoding an image, the method comprising:

outputting a first candidate list of the current block using the motion information of a neighboring block of the current block to output a first candidate list for the current block;

deriving a second list of candidates for the current block using the previously recovered motion information;

outputting a third list of candidates using the first list of candidates and the second list of candidates; And

generating a prediction block for the current block using the third candidate list.

2. The method of claim 1, wherein the adjacent block includes at least one of a spatially adjacent block and a temporally adjacent block.

3. The method according to claim 1, wherein producing a second list of candidates includes:

adding previously recovered motion information to the second list of candidates,

wherein the previously recovered motion information is motion information of a coding block decoded immediately before decoding the current block.

4. The method according to claim 3, additionally containing:

initializing the second list of candidates,

wherein the second candidate list is initialized based on a coding tree unit (CTU) line boundary.

5. The method according to claim 3, additionally containing:

when the number of pieces of motion information included in the second candidate list is a predetermined value, removing the motion information included in the second candidate list from the first among the pieces of motion information included in the second candidate list.

6. The method of claim 5, wherein the previously recovered motion information is added in the following order of motion information included in the second candidate list last.

7. The method according to claim 3, additionally containing:

when the same motion information as the previously recovered motion information is already included in the second candidate list, removing the same motion information from the second candidate list.

8. The method of claim 7, wherein the previously recovered motion information is added in the following order of motion information included in the second candidate list last.

9. The method according to claim 3, wherein the maximum number of pieces of motion information that can be included in the second candidate list is predetermined.

10. The method according to claim 1, additionally containing:

deriving average motion information by calculating an average value of a plurality of pieces of motion information included in the third candidate list; And

adding average motion information to the third list of candidates.

11. The method according to claim 1, additionally containing:

dividing the current block into the first area and the second area,

wherein generating a prediction block for the current block includes generating prediction blocks for the first region and the second region using the third candidate list.

12. The method of claim 1, wherein the third candidate list is used to generate an intra-block copy (IBC) prediction block for the current block.

13. A method for encoding an image, the method comprising:

determining motion information of the current block;

outputting a first list of candidates for the current block using motion information of an adjacent block of the current block;

outputting a second list of candidates for the current block using the previously encoded motion information; And

outputting a third list of candidates using the first list of candidates and the second list of candidates.

14. The method according to claim 13, wherein producing a second list of candidates includes:

adding previously encoded motion information to the second candidate list,

wherein the previously encoded motion information is the motion information of an encoding block encoded immediately before encoding the current block.

15. The method according to claim 14, additionally containing:

initializing the second list of candidates,

wherein the second candidate list is initialized based on a coding tree unit (CTU) line boundary.

16. The method according to claim 14, additionally containing:

when the number of pieces of motion information included in the second candidate list is a predetermined value, removing the motion information included in the second candidate list from the first among the pieces of motion information included in the second candidate list.

17. The method according to claim 14, additionally containing:

when the same motion information as the previously encoded motion information is already included in the second candidate list, removing the same motion information from the second candidate list.

18. The method of claim 14, wherein the number of pieces of motion information that can be included in the second candidate list is predetermined.

19. A computer-readable recording medium storing a bit stream that is received by a picture decoding apparatus and used to reconstruct the current block included in the current picture,

wherein the bit stream includes prediction information of the current block;

the prediction information is used to output a first candidate list for the current block using motion information of a neighboring block of the current block;

the prediction information is used to derive a second candidate list for the current block using the previously recovered motion information; And

the first candidate list and the second candidate list are used to output a third candidate list used to generate a prediction block for the current block.

1. A method for decoding an image, the method comprising:

outputting a first list of candidates of the current block using motion information of an adjacent block of the current block;

deriving a second list of candidates for the current block using the previously recovered motion information;

outputting a third list of candidates using the first list of candidates and the second list of candidates; And

generating a prediction block for the current block using the third list of candidates, while

when there is a candidate in the second candidate list that is equal to the previously recovered motion information, that candidate is removed from the second candidate list, and the previously recovered motion information is added to the last index position of the second candidate list.

2. The method of claim 1, wherein the adjacent block includes a spatially adjacent block and a temporally adjacent block.

3. The method according to paragraph 1, in which the output of the second list of candidates includes:

adding previously recovered motion information to the second list of candidates,

wherein the previously recovered motion information is motion information of a coding block decoded immediately before decoding the current block.

4. The method according to paragraph 3, additionally containing:

initializing the second list of candidates,

wherein the second candidate list is initialized based on a coding tree unit (CTU) line boundary.

5. The method according to paragraph 3, additionally containing:

when the number of motion information pieces included in the second candidate list is a predetermined value, removing the motion information included in the second candidate list first among the motion information pieces included in the second candidate list.

6. The method of claim 3, wherein the maximum number of pieces of motion information that can be included in the second candidate list is predetermined.

7. The method according to paragraph 1, additionally containing:

deriving average motion information by calculating an average value of a plurality of pieces of motion information included in the third candidate list; And

adding average motion information to the third list of candidates.

8. The method according to paragraph 1, in which

the first prediction block and the second prediction block are generated using the third candidate list,

the prediction block for the current block is generated based on the first prediction block and the second prediction block,

a first partial prediction block region for the current block is generated based on the first prediction block, and

a second partial prediction block region for the current block is generated based on the second prediction block.

9. The method of claim 1, wherein the third candidate list is used to generate an intra-block copy (IBC) prediction block for the current block.

10. A method for encoding an image, the method comprising:

determining motion information of the current block;

outputting a first list of candidates for the current block using motion information of the adjacent block relative to the current block;

outputting a second list of candidates for the current block using the previously encoded motion information; And

outputting a third list of candidates using the first list of candidates and the second list of candidates, wherein

when there is a candidate in the second candidate list that is equal to the previously recovered motion information, that candidate is removed from the second candidate list, and the previously recovered motion information is added to the last index position of the second candidate list.

11. The method according to paragraph 10, in which the output of the second list of candidates includes:

adding previously encoded motion information to the second candidate list,

wherein the previously encoded motion information is the motion information of an encoding block encoded immediately before encoding the current block.

12. The method according to paragraph 11, additionally containing:

initializing the second list of candidates,

wherein the second candidate list is initialized based on a coding tree unit (CTU) line boundary.

13. A computer-readable medium storing a bitstream including computer-executable code, wherein the computer-executable code, when executed by a processor of a video decoding device, causes the processor to perform steps comprising:

decoding prediction information in computer-executable code;

outputting a first list of candidates of the current block using motion information of an adjacent block of the current block;

deriving a second list of candidates for the current block using the previously recovered motion information;

outputting a third list of candidates using the first list of candidates and the second list of candidates; And

generating a prediction block for the current block using the third list of candidates, while

the prediction information is used to derive the first candidate list and the second candidate list, and

when there is a candidate in the second candidate list that is equal to the previously recovered motion information, that candidate is removed from the second candidate list, and the previously recovered motion information is added to the last index position of the second candidate list.

Claims (67)

1. A method for decoding an image, the method comprising: outputting a list of merge candidates for the current block using information of a first neighboring block of said current block; And generating a final prediction block for the current block using the list of merge candidates, while the final prediction block for the current block is generated based on the first prediction block and the second prediction block, the first prediction block is generated based on the first candidate from the list of merge candidates, the second prediction block is generated based on the second candidate from the list of merge candidates, the first prediction block is used to determine the first partial region of the final prediction block for the current block, a second prediction block is used to determine a second partial region of the final prediction block for the current block, and the final prediction block for the current block is generated using a weighted sum using the first prediction block and the second prediction block, different weights are applied to multiple regions in the first prediction block to generate the final prediction block, respectively, and different weights are applied to a plurality of regions in the second prediction block to generate the final prediction block, respectively. 2. The method according to claim 1, additionally containing: deriving a first list of candidates for the current block by using motion information of a second adjacent block of said current block; And outputting a second list of candidates for the current block using previously recovered motion information, wherein the list of merge candidates for the current block is output using the first list of candidates and the second list of candidates, wherein the second adjacent block includes at least one of a spatially adjacent block and a temporally adjacent block. 3. The method according to claim 2, in which the output of the second list of candidates includes adding previously recovered motion information to the second list of candidates, wherein the previously recovered motion information is motion information of a coding block decoded immediately before decoding the current block. 4. The method according to claim 3, additionally containing initializing the second list of candidates, wherein the second candidate list is initialized based on a coding tree unit (CTU) line boundary. 5. The method according to claim 3, additionally containing, when the number of pieces of motion information included in the second candidate list is a predetermined value, removing the motion information included in the second candidate list from the first among the pieces of motion information included in the second candidate list. 6. The method according to claim 3, additionally containing, when the same motion information as the previously recovered motion information is already included in the second candidate list, removing the same motion information from the second candidate list. 7. The method of claim 3, wherein the maximum number of pieces of motion information that can be included in the second candidate list is predetermined. 8. The method according to claim 1, additionally containing: deriving average motion information by calculating the average value of a plurality of pieces of motion information included in the fusion candidate list; And adding average motion information to the merge candidate list. 9. The method according to claim 1, additionally containing dividing the current block into a first region and a second region, wherein generating a prediction block for the current block includes generating prediction blocks for the first region and the second region using the list of merge candidates. 10. The method of claim 1, wherein the list of merge candidates is used to generate an intra-block copy (IBC) prediction block for the current block. 11. A method for encoding an image, the method comprising: outputting a list of merge candidates for the current block using information of a first neighboring block of said current block; And generating a final prediction block for the current block using the list of merge candidates, while the final prediction block for the current block is generated based on the first prediction block and the second prediction block, the first prediction block is generated based on the first candidate from the list of merge candidates, the second prediction block is generated based on the second candidate from the list of merge candidates, the first prediction block is used to determine the first partial region of the final prediction block for the current block, a second prediction block is used to determine a second partial region of the final prediction block for the current block, and the final prediction block for the current block is generated using a weighted sum using the first prediction block and the second prediction block, different weights are applied to multiple regions in the first prediction block to generate the final prediction block, respectively, and different weights are applied to a plurality of regions in the second prediction block to generate the final prediction block, respectively. 12. The method according to claim 11, additionally containing: deriving a first list of candidates for the current block by using motion information of a second adjacent block of said current block; And outputting a second list of candidates for the current block using previously recovered motion information, wherein the list of merge candidates for the current block is output using the first list of candidates and the second list of candidates, in which output of the second list of candidates includes adding previously encoded motion information to the second candidate list, wherein the previously encoded motion information is the motion information of an encoding block encoded immediately before encoding the current block. 13. The method according to claim 12, additionally containing initializing the second list of candidates, wherein the second candidate list is initialized based on a coding tree unit (CTU) line boundary. 14. A computer-readable medium storing a bit stream including computer executable code, wherein the computer executable code, when executed, causes a video decoding device to perform the following steps comprising: decoding prediction information about the prediction for the current block in the computer-executable code; outputting a list of merge candidates for the current block using the information of the first neighboring block of the current block and the prediction information; generating a prediction block for the current block using the list of merge candidates; And generating a final prediction block for the current block using the list of merge candidates, while the final prediction block for the current block is generated based on the first prediction block and the second prediction block, the first prediction block is generated based on the first candidate from the list of merge candidates, the second prediction block is generated based on the second candidate from the list of merge candidates, the first prediction block is used to determine the first partial region of the final prediction block for the current block, a second prediction block is used to determine a second partial region of the final prediction block for the current block, and the final prediction block for the current block is generated using a weighted sum using the first prediction block and the second prediction block, different weights are applied to multiple regions in the first prediction block to generate the final prediction block, respectively, and different weights are applied to a plurality of regions in the second prediction block to generate the final prediction block, respectively.