KR101891192B1 - Method and Apparatus for image encoding - Google Patents

Abstract

The present invention relates to a video coding method, and more particularly, to a video coding method and a motion vector coding method, Selecting a candidate motion vector in the search area; And estimating a cost for the selected candidate motion vector, thereby improving the compression efficiency by more accurately performing motion vector prediction of the enhancement layer.

Description

[0001] The present invention relates to a method and apparatus for image encoding,

The present invention relates to a video encoding method and apparatus.

H.264 / AVC is a video compression standard technology with high compression efficiency. In H.264 / AVC, the original signal can be predictively encoded through an intra prediction technique for eliminating correlation in an image and an inter prediction technique for eliminating correlation between images. The H.264 / AVC encoder performs discrete cosine transform (DCT) encoding and quantization on the difference value between the original signal and the prediction signal. The quantized signal is then entropy encoded after being aligned by a zigzag scanning method.

Recently, ITU-T VCEG (Video Coding Experts Group) and ISO / IEC MPEG (Moving Picture Experts Group) have formed Joint Collaborative Team on Video Coding (JCT-VC) And it is known that the compression efficiency is improved by about 40% or more as compared with the existing standard H.264 / AVC. H.264 / AVC and HEVC are basically block-based image coders. However, unlike H.264 / AVC, which encodes MB (Macro Block) units with a fixed size of 16x16, Encoding according to a CU (Coding Unit) basis having various sizes ranging from a maximum 64x64 size to an 8x8 size.

Major technologies related to motion vectors implemented in HEVC are Merge technology and Advanced Motion Vector Prediction (AMVP) technology.

The Merge technique is a technique for inserting appropriate peripheral motion vectors into a Merge candidate list, and then checking a list in the candidate group to decide whether to merge with the corresponding motion vector. AMVP technology is a multi-candidate based motion vector determination technique, A motion vector candidate group is generated using appropriate surrounding motion vectors, an optimal motion vector is selected in a candidate list included in the candidate group, and an index indicating a candidate group and an optimal motion vector are simultaneously selected.

Since the HEVC achieves a high coding efficiency through the above two techniques, the above two techniques are also used in the HEVC (Spatial Scalable Coding) method based on the High Efficiency Video Coding (HEVC).

At this time, previous researches have been conducted to achieve a higher compression efficiency by using information of a base layer in generating a Merge candidate list and an AMVP candidate list.

However, in the conventional HEVC-based spatial layer coding process, the information of the base layer used for predicting the motion vector of the enhancement layer is limited to the motion vector.

When the LCU of the base layer is divided into lower CUs and has NxN type PUs, the motion vector of the base layer used for determining the optimal motion vector mv * of the enhancement layer LCU is maximum 4 (mv0, mv1, mv2 , mv3).

Further, conventionally, since a motion vector of a base layer is scaled and used as it is, there is no method for selecting an optimal motion vector among the four motion vectors of the base layer.

In addition, when four motion vectors of the base layer have different values, one of four motion vectors is selected and used as it is.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and it is an object of the present invention to provide a video encoding method and apparatus for improving compression efficiency in encoding.

Another object of the present invention is to provide an image encoding method and apparatus for predicting an appropriate motion vector for a PU of an enhancement layer.

According to an aspect of the present invention, there is provided an image encoding method including the steps of setting a motion vector of a base layer and a predetermined range of a motion vector of the base layer as a search region, Selecting a candidate motion vector in the candidate motion vector, and estimating a cost for the selected candidate motion vector.

Here, it is preferable that the search area is determined by a sum of a motion vector of the base layer and a search motion vector.

In addition, the search motion vector may be composed of any combination of two sets of {-a, 0, a} (where a is a natural number).

The cost prediction may be determined based on a distance between the motion vector of the base layer and the search motion vector.

Here, it is preferable to use the distortion prediction function having a characteristic that the distortion value increases as the distance between the motion vector of the base layer and the search motion vector increases, but the rate of increase of the distortion value decreases Do.

It is preferable that the cost prediction step uses a cost prediction function including a bit amount generated when the search motion vector and a prediction motion vector of the enhancement layer are coded and the distortion prediction function.

In addition, it is preferable that, in the step of predicting the cost, each cost is calculated with the cost prediction function for a plurality of motion vectors of the base layer, and then the final predicted cost, which is a sum of the costs, is calculated.

The image encoding method may further include determining the search motion vector having the smallest final prediction cost as a final candidate motion vector in the step of estimating the cost.

Meanwhile, the final candidate motion vector may be scaled according to the spatial expansion ratio.

In addition, the final candidate motion vector is preferably added to a merge candidate list or an Advance Motion Vector Prediction candidate list.

According to the present invention, an accurate motion vector for an enhancement layer can be obtained by using extended base layer information such as a motion vector and a residual signal, thereby improving encoding efficiency.

In addition, when motion vectors of a plurality of different base layers are present, it is possible to determine a predicted motion vector by searching a certain range around the motion vector, thereby improving the coding efficiency of the enhancement layer.

1 schematically shows an example of the structure of an image encoding apparatus.
2 schematically shows an example of the structure of a video decoding apparatus.
3 shows an example of a picture to be coded / decoded and a reference picture.
4 is a flowchart illustrating an image encoding method according to an embodiment of the present invention.
FIG. 5 illustrates a method of setting a motion vector search region included in the image encoding method according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Hereinafter, the same reference numerals will be used for the same constituent elements in the drawings, and redundant explanations for the same constituent elements will be omitted.

1 schematically shows an example of the structure of an image encoding apparatus.

1, the image encoding apparatus 100 includes a motion prediction unit 111, a motion compensation unit 112, an intra prediction unit 120, a switch 115, a subtractor 125, a transform unit 130, A quantization unit 140, an entropy encoding unit 150, an inverse quantization unit 160, an inverse transformation unit 170, an adder 175, a filter unit 180, and a reference image buffer 190.

The image encoding apparatus 100 performs encoding in an intra mode or an inter mode with respect to an input image and outputs a bit stream. In the embodiment of the present invention, intra prediction can be used in the same way as inter prediction, and inter prediction can be used in the same meaning as inter prediction. The intra prediction method and the inter prediction method may be selectively used for the prediction unit in order to determine an optimal prediction method for the prediction unit. The image encoding apparatus 100 generates a prediction block for the original block of the input image, and then encodes the difference between the original block and the prediction block.

In the intra prediction mode, the intra prediction unit 120 (or the intra prediction unit can also be used as a term having the same meaning) performs spatial prediction using the pixel values of the already coded blocks around the current block And generates a prediction block.

In the case of the inter-picture prediction mode, the motion prediction unit 111 finds a motion vector by searching an area of the reference picture stored in the reference picture buffer 190, which is best matched with the input block, in the motion prediction process. The motion compensation unit 112 generates a prediction block by performing motion compensation using a motion vector.

The subtracter 125 generates a residual block by a difference between the input block and the generated prediction block. The transforming unit 130 performs a transform on the residual block to output a transform coefficient. The quantization unit 140 quantizes the input transform coefficient according to the quantization parameter and outputs a quantized coefficient. The entropy encoding unit 150 entropy-codes the input quantized coefficients according to a probability distribution to output a bit stream.

Since the HEVC performs inter prediction coding, i.e., inter prediction coding, the currently encoded image needs to be decoded and stored for use as a reference image. Accordingly, the quantized coefficients are inversely quantized in the inverse quantization unit 160 and inversely transformed in the inverse transformation unit 170. The inverse quantized and inverse transformed coefficients are added to the prediction block through the adder 175 and a reconstruction block is generated.

The restoration block passes through the filter unit 180 and the filter unit 180 applies at least one of a deblocking filter, a sample adaptive offset (SAO), and an adaptive loop filter (ALF) can do. The filter unit 180 may be referred to as an adaptive in-loop filter. The deblocking filter can remove block distortion occurring at the boundary between the blocks. The SAO may add a proper offset value to the pixel value to compensate for coding errors. The ALF may perform filtering based on a comparison between the reconstructed image and the original image, and may be performed only when high efficiency is applied. The restoration block having passed through the filter unit 180 is stored in the reference image buffer 190.

2 schematically shows an example of the structure of a video decoding apparatus.

2, the image decoding apparatus 200 includes an entropy decoding unit 210, an inverse quantization unit 220, an inverse transform unit 230, an intra prediction unit 240, a motion compensation unit 250, 260 and a reference image buffer 270.

The video decoding apparatus 200 receives the bit stream output from the encoder and decodes the video stream into an intra mode or an inter mode, and outputs a reconstructed video, i.e., a reconstructed video. In the intra mode, a prediction block is generated using an intra prediction mode, and a prediction block is generated using an inter prediction method in an inter mode. The video decoding apparatus 200 obtains a residual block from the input bitstream, generates a prediction block, and then adds the residual block and the prediction block to generate a reconstructed block, that is, a reconstruction block.

The entropy decoding unit 210 entropy-decodes the input bitstream according to a probability distribution and outputs a quantized coefficient. The quantized coefficients are inversely quantized in the inverse quantization unit 220 and inversely transformed in the inverse transformation unit 230. As a result of inverse quantization / inverse transformation of the quantized coefficients, a residual block is generated.

In the intra-picture prediction mode, the intra-prediction unit 240 (or the inter-picture prediction unit) performs spatial prediction using the pixel values of the already coded blocks around the current block to generate a prediction block.

In the inter-view prediction mode, the motion compensation unit 250 generates a prediction block by performing motion compensation using a motion vector and a reference image stored in the reference image buffer 270.

The residual block and the prediction block are added through the adder 255, and the added block is passed through the filter unit 260. [ The filter unit 260 may apply at least one of a deblocking filter, SAO, and ALF to a restoration block or a restored picture. The filter unit 260 outputs a reconstructed image, that is, a reconstructed image. The restored image is stored in the reference image buffer 270 and can be used for inter-view prediction.

Methods for improving the prediction performance of the encoding / decoding apparatus include a method of increasing the accuracy of the interpolation image and a method of predicting the difference signal. Here, the difference signal is a signal indicating the difference between the original image and the predicted image.

In the present invention, the term " difference signal " may be replaced by a " difference signal ", " residual block ", or " difference block " depending on the context. Those skilled in the art may influence the idea You will be able to distinguish this within the scope of not giving.

In the embodiment of the present invention, a coding unit (CU) is used as a coding unit for convenience of explanation, but it may be a unit for performing not only coding but also decoding. Hereinafter, the image encoding method described in the embodiment of the present invention can be implemented in accordance with the functions of the respective modules described in FIGS. 1 and 2, and such an encoder and a decoder are included in the scope of the present invention. That is, the image encoding / decoding method to be described later in the embodiment of the present invention can be performed in each component included in the image encoder and the image decoder described in FIG. 1 and FIG. The meaning of the constituent part may include not only a hardware meaning but also a software processing unit which can be performed through an algorithm.

In the image encoding / decoding according to the embodiments of FIGS. 1 and 2, a CU structure in which a single-sized macro block is expanded to various sizes may be defined for efficient encoding of an image. The CU is a unit in which a video is encoded by a video encoder and can be hierarchically divided with depth information based on a quad tree structure. CUs can have various sizes such as 8 × 8, 16 × 16, 32 × 32, and 64 × 64. The largest CU is called the LCU (Largest Coding Unit), and the smallest CU is called the SCU (Smallest Coding Unit). All CUs except SCUs are assigned split_flag information and indicate whether or not the corresponding CU is divided according to the value of split_flag. The encoder may adjust the size of the LCU in the encoding process according to various video signal characteristics. The CU can be divided into PU (Prediction Unit) for use in intra or inter prediction, and can be divided into TU (Transform Unit) for conversion and quantization. Hereinafter, a block means a unit of encoding / decoding. In the encoding / decoding process, the image is divided into a predetermined size and then encoded / decoded. Thus, a block may also be referred to as CU, PU, TU, etc., and one block may be divided into sub-blocks of smaller size.

Here, PU denotes a basic unit of prediction and / or motion compensation performance. A PU can be divided into a plurality of partitions, and each partition is called a PU partition (prediction unit partition). When the PU is divided into a plurality of partitions, the PU partitions can be a basic unit of performing prediction and / or motion compensation. Hereinafter, in the embodiment of the present invention, the PU may mean a PU partition.

Meanwhile, HEVC (High Efficiency Video Coding) uses a motion vector prediction method based on Advanced Motion Vector Prediction (AMVP).

In the motion vector prediction method based on the improved motion vector prediction method, not only the motion vector (MV) of the reconstructed block located around the current block to be encoded / decoded but also the motion vector of the current block to be encoded / decoded in the reference picture A motion vector of a block existing in a position or a corresponding position can be used. At this time. A block located at the same position or a position corresponding to a spatially corresponding position in the reference picture as a block to be coded / decoded is referred to as a collocated block, a motion vector of an equal position block is referred to as a collocated motion vector or a temporal motion It is called a temporal motion vector. However, the collocated block may be a block existing in the same position as the block to be encoded / decoded, that is, a block existing at the corresponding position, have.

In the motion information merge method, motion information is inferred not only from neighboring reconstructed blocks but also from equivalent position blocks, and uses the motion information as motion information of a current block to be encoded / decoded. At this time, the motion information includes inter-prediction mode information indicating a reference picture index, a motion vector, a uni-direction or a bi-direction necessary for inter prediction, a reference picture list ), Prediction mode information about whether the image is coded in the intra-prediction mode or the inter-prediction mode, and the like.

The predicted motion vector predicted in the current block to be coded / decoded includes not only motion vectors of neighboring blocks spatially adjacent to the current block to be coded / decoded, but also motion vectors of neighboring blocks that are temporally adjacent to the current block to be coded / Lt; / RTI >

3 shows an example of a picture to be coded / decoded and a reference picture.

Referring to FIG. 3, a block X represents a current block to be encoded / decoded in a picture 310 to be encoded / decoded. Block A, block B, block C, block D and block E are located around a block to be encoded / Lt; / RTI > A block T in the reference picture 320 indicates an equivalent position block existing at a position corresponding to the block to be encoded / decoded.

4 is a flowchart illustrating an image encoding method according to an embodiment of the present invention.

Referring to FIG. 4, a method of encoding an image according to an exemplary embodiment of the present invention includes selecting a candidate motion vector by setting a motion vector of a base layer and a surrounding range of a motion vector of the base layer as a search region do.

First, a motion vector mv _i and a residual signal c _i are extracted from a bitstream of a base layer corresponding to a PU of the enhancement layer (S110). Here, the motion vector (mv _i , i = 0, ..., K) of the base layer is derived from the PU of the base layer corresponding to the PU of the enhancement layer. The residual signal (c _i ) is calculated as the sum of the absolute values of the motion estimation errors when the corresponding mvi is used in the base layer PU.

Next, a certain range P around the motion vector mv _i of the base layer is set as the motion vector search area S (S120). One embodiment for setting the search area will be described later.

Then, it is checked whether all the motion vectors in the motion vector search area are searched (S130).

If all the motion vectors are not found as a result of the check, a search motion vector is selected in the motion vector search area (S140).

Thereafter, the cost is predicted using the motion vector cost prediction function for the selected search motion vector (S150).

Next, the search motion vector and the predicted cost are stored in the memory (S160), and the step of checking again whether all the motion vectors in the motion vector search area are searched is recursively performed.

At this time, if all the motion vectors in the motion vector search area are searched, the minimum value among the predicted costs stored in the memory is searched for and the corresponding search motion vector is selected as the final candidate motion vector (S170).

Thereafter, the final candidate motion vector is scaled according to the spatial expansion ratio (S180).

Finally, the scaled final candidate motion vector is added to the merge candidate list for the PU of the enhancement layer and the AMVP candidate list (S190).

FIG. 5 illustrates a method of setting a motion vector search region included in the image encoding method according to an embodiment of the present invention.

Referring to FIG. 5, a predetermined range P of a total of 9 pixels is set in the horizontal / vertical direction around the motion vector mv _i of the base layer, and is set to the motion vector search area S. Here, the search area S is determined by the sum of the base layer motion vector mv _i and the search motion vector mv _r .

In Figure 5, the search motion vector (mv _r) a set consisting of, but composed of {4, 0, 4}, which is one embodiment of the example, when the normally appear, {-a, 0, a} (where, and a is a natural number).

The P and S can be expressed by the following equations.

[Equation 1]

_{P = {mv r = (x} , y): x, y ∈ {-4, 0, 4}, r = 1, ..., 9}

&Quot; (2) "

S = {mv _s = mv _i + mv _r : i = 1, ..., 3, r = 1, ..., 9}

Next, by substituting the selected search motion vector (mv _s), a motion vector prediction cost function (equation 3) predicts the cost of using the motion vector.

Here, since the cost prediction function is a motion vector obtained as a result of the rate-distortion optimization of the base layer motion vector mv _i , as the search motion vector mv = mv _s is farther from the motion vector of the base layer, Based on the assumption that it increases.

Accordingly, in Equation (3), a function capable of measuring the distance (l) between two vectors is defined. In addition to Equation (3), various methods for measuring the distance between two vectors can be used in the present invention.

&Quot; (3) "

here,

(1) between the two vectors obtained in Equation (3) and a distortion prediction function of Equation (4) according to the present invention. The proposed distortion prediction function has the property that the distortion value increases as the distance (l) increases, but the rate of increase decreases. It also has a and b function parameters, which are the results obtained through various experimental images in advance.

&Quot; (4) "

The motion vector of the base layer is calculated by summing the bit amount R (mv) generated when the PMV (Predictive Motion Vector) and the search motion vector (mv) of the enhancement layer PU are coded with the Lagrangian exponent, mvi) can be predicted as shown in Equation (5). Here, the method of obtaining PMV and R (mv) is the same as the method provided by HM, the HEVC reference software.

&Quot; (5) "

K motion vector of the base layer corresponding to the enhancement layer _{PU (mv i, i = 0} , ..., K) obtained by adding together that the cost using the cost function of the prediction equation (5) with respect to as in the equation (6) The final predicted cost of the search motion vector (mv = mv _s ) is obtained.

&Quot; (6) "

The motion vector having the smallest cost among the search motion vectors is determined as the final candidate motion vector mv +.

Scales the final candidate motion vector mv + according to the spatial expansion ratio, and adds the scaled motion vector mv * to the Merge candidate list and the AMVP candidate list for the enhancement layer PU.

As described above, in the HEVC-based spatial layer coding process, the base layer information used for predicting the motion vector of the enhancement layer is limited to the motion vector. However, the present invention extends the motion vector and residual signal, The motion vector of the motion vector is obtained.

When the LCU of the base layer is divided into lower CUs and has NxN type PUs, the motion vector of the base layer used for determining the optimal motion vector mv * of the enhancement layer LCU is maximum 4 (mv0, mv1, mv2 , mv3). In this case, the optimal motion vector among the four motion vectors of the base layer can be selected. That is, it is possible to correctly perform the motion vector prediction of the enhancement layer among the motion vectors of the base layer and the motion vectors of a certain range around the four motion vectors.

100: image encoding device 111: motion prediction unit
112: motion compensation unit 120: intra prediction unit
125: subtracter 130:
140: quantization unit 150: entropy coding unit

Claims (20)

A video encoding method supporting a plurality of layers,
(I) a motion vector of the base layer corresponding to the enhancement layer, and (ii) a residual signal calculated using the motion prediction error, extracts a motion vector of the base layer based on the residual signal, Setting a search area determined by a sum of a motion vector of a layer and a search motion vector;
Selecting a candidate motion vector in the search area; And
And estimating a cost of the selected candidate motion vector by determining a distance between the motion vector of the base layer and the search motion vector. delete The method according to claim 1,
Wherein the search motion vector is composed of any two combinations of a set {-a, 0, a} (where a is a natural number). delete The method according to claim 1,
Wherein the predicting the cost uses a distortion prediction function having a characteristic that the distortion value increases as the distance between the motion vector of the base layer and the search motion vector increases but the rate of increase of the distortion value decreases, . The method of claim 5,
Wherein the predicting the cost uses a cost prediction function including a bit amount generated when the search motion vector and a prediction motion vector of the enhancement layer are coded and the distortion prediction function. The method of claim 6,
Wherein the cost prediction step calculates each cost by using the cost prediction function for a plurality of motion vectors of the base layer, and then calculates a final prediction cost by summing the costs. The method of claim 7,
The image coding method according to claim 1, further comprising the step of determining the search motion vector having the smallest final prediction cost as a final candidate motion vector in the step of predicting the cost. The method of claim 8,
Wherein the final candidate motion vector is scaled according to a spatial expansion ratio. The method of claim 9,
Wherein the final candidate motion vector is added to a Merge candidate list or an Advance Motion Vector Prediction candidate list. A processor; And
And a memory coupled to the processor and storing information for driving the processor,
The processor extracts a residual signal calculated using (i) the motion vector of the base layer corresponding to the enhancement layer and (ii) the motion prediction error, and calculates a residual range of a motion vector of the base layer based on the residual signal A motion vector of the base layer and a search motion vector, a candidate motion vector is selected in the search region, and a distance between the motion vector of the base layer and the search motion vector And predict a cost for the selected candidate motion vector. delete The method of claim 11,
Wherein the search motion vector is composed of any two combinations in a set {-a, 0, a} (where a is a natural number). delete The method of claim 11,
Wherein the prediction of the cost is configured to use a distortion prediction function having a characteristic that the distortion value increases as the distance between the motion vector of the base layer and the search motion vector increases but the rate of increase of the distortion value decreases. . 16. The method of claim 15,
Wherein the cost prediction unit is configured to use a cost prediction function including a bit amount generated when the search motion vector and a prediction motion vector of the enhancement layer are coded and the distortion prediction function. 18. The method of claim 16,
Wherein the prediction of the cost is configured to calculate each cost with the cost prediction function for a plurality of motion vectors of the base layer and then calculate a final predicted cost by summing the costs. 18. The method of claim 17,
Wherein the image encoding apparatus is configured to determine the search motion vector having the smallest final prediction cost as a final candidate motion vector in the step of predicting the cost. 19. The method of claim 18,
Wherein the final candidate motion vector is configured to be scaled according to a spatial expansion ratio. The method of claim 19,
Wherein the final candidate motion vector is configured to be added to a Merge candidate list or an Advance Motion Vector Prediction candidate list.

Images