CN102215395B - Video coding and decoding method and device - Google Patents

Abstract

An encoding method, comprising: selecting an irregular division mode among a plurality of preset irregular division modes to irregularly divide an image block to be encoded to obtain at least two different sub-image blocks, and according to the sub-image blocks Obtain the reference image block of the sub-image block according to the pixel value of the sub-image block; obtain the motion vector of the sub-image block of the image block to be encoded according to the reference image block; obtain the motion vector of the sub-image block of the image block to be encoded; Obtain the reference motion vector of the sub-image block from the encoded peripheral block; obtain the motion vector difference value according to the motion vector and the reference motion vector; write the motion vector difference value into the code stream and send it to the decoder end. The embodiment of the present invention also provides an encoding device for executing the encoding method.

Description

A video encoding and decoding method and device

technical field

The invention relates to the field of video codec, in particular to a video codec method and device for dividing irregular sub-image blocks.

Background technique

In the existing video coding and decoding technology, the residual matrix is generally obtained by making a difference between the pixel value of the image block of the current frame image and the pixel value of the reference image block, and then the obtained residual matrix is quantized, transformed and After entropy encoding, it is written into the code stream and sent to the decoding end together with the motion vector information. The decoding end reads the bit stream from the obtained code stream, and then undergoes inverse quantization and inverse transformation to obtain the corresponding residual, and then obtains the reference image block of the current image block according to the motion vector information in the code stream, and converts The residual of the current image block is added to the pixel value of the reference image block to obtain the pixel value of the current image block.

In the encoding and decoding process of the prior art, both the calculation of the residual and the search of the reference image block are carried out in units of the sub-image blocks of the image block, and there are corresponding methods for dividing Regular division and irregular division technology of sub-image blocks. Regular division is to divide the image block into identical square sub-image blocks. Its advantage is that the acquisition of various parameters is relatively simple, such as the reference motion vector used to predict the position of the reference image, but its disadvantage is that the exact same square sub-image blocks The shape of the image is often very different from the outline of the object in the image, which increases the difficulty of pixel prediction in encoding and decoding. Correspondingly, the irregular division technology is to divide the image block into rectangles, triangles, even pentagons and irregular quadrilaterals of different shapes, so as to meet the requirements of different types of images, which can greatly solve the pixel prediction problem of regular division. However, correspondingly, due to the irregular shape, the non-regular division technology also faces the problem of reference motion vector parameters used to calculate the motion vector of the image block.

Therefore, it is necessary to provide an encoding and decoding method and device for simply obtaining reference motion vectors of sub-image blocks in irregular division.

Contents of the invention

An embodiment of the present invention provides an encoding method, comprising: selecting an irregular division mode among a plurality of preset irregular division modes to divide an image block to be encoded irregularly to obtain at least two different sub-image blocks, Obtain the reference image block of the sub-image block according to the pixel value of the sub-image block; obtain the motion vector of the sub-image block of the image block to be encoded according to the reference image block; obtain the motion vector of the sub-image block of the image block to be encoded according to the image block to be encoded The coded peripheral block of the sub-image block and the irregular division method of the image block obtain the reference motion vector of the sub-image block; according to the motion vector and the reference motion vector, a motion vector difference is obtained; the motion vector The vector difference is written into the code stream and sent to the decoder. The embodiment of the present invention also provides an encoding device for executing the encoding method.

The embodiment of the present invention also provides a corresponding decoding method, including obtaining the irregular division mode of the image block from the code stream, and obtaining the motion vector difference value of the sub-image block of the image block to be decoded; from the image to be decoded According to the irregular division method of the decoded peripheral blocks of the sub-image block and the image block to be decoded, the reference motion vector of the sub-image block is obtained; according to the reference motion vector of the sub-image block and the difference value of the motion vector, the The motion vector of the sub-image block; obtain the reference image block of the sub-image block according to the motion vector, and decode the sub-image block according to the reference sub-image block, the embodiment of the present invention also provides a A decoding device for performing the decoding method.

The encoding and decoding method and device provided by the embodiment of the present invention can simplify the acquisition and calculation of the reference motion vector under certain conditions, save the resources of the encoding and decoding system, and improve the efficiency.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. Those of ordinary skill in the art can also obtain other drawings based on these drawings without any creative effort.

FIG. 1 is a schematic diagram of an encoding method provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram of an irregular division method provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram showing the distribution of surrounding blocks of a block to be encoded in an embodiment of the present invention;

FIG. 4 is a schematic diagram of an encoding device provided by an embodiment of the present invention;

FIG. 5 is a schematic diagram of a decoding method provided by an embodiment of the present invention;

FIG. 6 is a schematic diagram of a decoding device provided by an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

Referring to Fig. 1, an embodiment of the present invention provides a video coding method, which includes:

Step 101: Select an irregular division method among the preset irregular division methods to divide the image block to be coded irregularly to obtain at least two different sub-image blocks, and obtain according to the pixel values of the sub-image blocks a reference image block of the sub-image block;

In the embodiment of the present invention, the irregular division of the image block to be coded includes four irregular division methods, please refer to Figures 2a to 2d, these four division methods are divided into The image block to be encoded is divided into a triangular sub-image block and a pentagonal sub-image block (respectively identified as Mode0 and Mode1), and the image block to be encoded is divided into two by vertical dividing lines and horizontal dividing lines. Rectangular sub-image blocks with inconsistent sizes (respectively identified as Mode2 and Mode3), where the parameter n is an irregularly divided position parameter, which is used to indicate the distance between the dividing line and the intersection point of the image block to be decoded.

In the above division scheme, when the division method shown in Fig. 2a and Fig. 2b is adopted for the image block, if the value of n is 0, the image block is divided into two triangles. Since the image blocks are composed of pixels, the 135-degree boundary line and the 45-degree boundary line corresponding to Mode0 and Mode1 refer to lines approximately 135-degree and 45-degree composed of pixels along the boundaries between sub-image blocks.

The selection of the irregular division method, the confirmation of the position parameter, and the confirmation of the reference image block in the embodiment of the present invention can be done by trying all the preset irregular division mode and the selectable position parameters, and The corresponding optimal reference image block is searched, and finally the results are compared to select the best solution. In the embodiment of the present invention, the optimal solution may be the division method and corresponding position parameters of the divided sub-image blocks closest to the searched reference image block pixels. Of course, this optimal solution may also be adjusted according to other requirements The plan is adjusted.

Step 102: Obtain a motion vector of a sub-image block of the image block to be encoded according to the reference image block;

Step 103: Obtain a reference motion vector of the sub-image block according to the coded surrounding blocks of the sub-image block of the image block to be encoded;

specific:

First, it is judged whether the encoded peripheral blocks meet the preset conditions. In the embodiment of the present invention, the preset conditions include:

The surrounding blocks adopt the same irregular division method as the image block to be coded;

The reference image block of the peripheral block and the reference image block of the sub-image block are in the same image.

Then, if there is only one peripheral block that satisfies the preset condition, the reference motion vector of the sub-image block is equal to the motion vector of the peripheral block that satisfies the condition.

If more than one of the coded peripheral blocks meets the above conditions, the reference motion vector of the sub-image block is equal to the motion of the peripheral block whose position parameter is closest to the image block to be coded among the peripheral blocks that meet the conditions vector. In the embodiment of the present invention, the closest may be that the difference between the location parameters is the smallest, or the absolute value of the difference between the location parameters is the smallest.

If the position parameters of multiple satisfying surrounding blocks are the same and are also close to the position parameters of the image block to be encoded, the reference motion vector of the sub-image block is the same as that of the multiple satisfying surrounding blocks. The average value of motion vectors, or the average value of motion vectors of surrounding blocks is rounded.

In the embodiment of the present invention, whether the location parameter is close or not can be judged numerically, such as a difference.

Step 104: Obtain a motion vector difference according to the motion vector and the reference motion vector;

Wherein, in the embodiment of the present invention, the motion vector difference obtained according to the appropriate amount of motion and the reference motion vector can be obtained by making a difference between the two, and in other optional embodiments, it can also be obtained by other Mathematical methods to establish the logical relationship between the motion vector difference and the motion vector and the reference motion vector, such as division, logarithmic calculation, etc.

Step 105: Write the motion vector difference into the code stream.

Using the encoding method provided by the embodiment of the present invention, under preset conditions, directly according to the motion vector of the sub-image block, the irregular division mode of the image block where the sub-image block is located, and the irregular division mode of the surrounding blocks of the sub-image block Obtaining the reference motion vector of the sub-image block can effectively save the number of coding bits and improve the coding efficiency compared with the complicated calculation in the prior art.

It should be pointed out that the image blocks, sub-image blocks and reference image blocks of surrounding blocks in the above solution are nouns defined for convenience of description. The image block in the above solutions refers to the basic unit of image coding, and one image block may contain one to several sub-image blocks. In the existing video coding and decoding technology, the image block can be a macroblock with a size of 16x16. At this time, the sub-image block is called a sub-block, and the size can be 16x16, 16x8, 8x16, 8x8, 8x4, 4x8, 4x4, or The irregularly divided blocks pointed out in the present invention; image blocks can also be super macroblocks with a size of 32x32. At this time, the sub-image block is a macroblock or a sub-block in a macroblock, and the size can be 32x32, 32x16, 16x32, 16x16, 16x8, 8x16, 8x8, 8x4, 4x8, 4x4, or the irregularly divided blocks pointed out by the present invention .

The surrounding blocks refer to image blocks or sub-image blocks adjacent to the current image block or the current sub-image block. The positional relationship between it and the current image block, or the current sub-image block can be adjacent, or pixel values at a fixed interval. The surrounding blocks of the current sub-image block may be image blocks or sub-image blocks.

The reference image blocks of the surrounding blocks may be the above-mentioned image blocks, or the above-mentioned sub-image blocks. Therefore, if the surrounding blocks of the current sub-image block are image blocks, the reference image blocks of the surrounding blocks of the current sub-image block can be image blocks; if the surrounding blocks of the current sub-image block are sub-image blocks, the current sub-image The reference image blocks of the peripheral blocks of the block may be sub-image blocks.

A further implementation will be given below.

In the following implementation schemes, since the surrounding blocks of the current image block or sub-image block can be image blocks or sub-image blocks, they are collectively referred to as image blocks for the convenience of description. The sizes of all image blocks may be inconsistent, and the shapes or division methods of all image blocks It can be inconsistent, and the size of the image block is determined according to the method in the prior art.

As shown in Figure 3, there are image blocks A, B, C, D and E. Where E is the current image block, A is the left adjacent block of E, B is the upper adjacent block of E, C is the upper right adjacent block of E, and D is the upper left adjacent block of E.

Let the motion vectors of image blocks A, B, C, and D be mvA, mvB, mvC, and mvD. Suppose the reference frame index values of image blocks A, B, C, D and E are refA, refB, refC, refD and refE.

The reference frame index value refers to the position in the video sequence of the image where the reference image block of the image block is located. When the reference frame index values of two image blocks are the same, the reference image blocks of the two image blocks are located in the same image.

It is assumed that the irregular block division methods of the image blocks A, B, C, D and E are ModeA, ModeB, ModeC, ModeD and ModeE.

If the image block A, or, B, or, C, or D is unavailable or does not use irregular block division, set the value of ModeA, or, ModeB, or, ModeC, or, ModeD to -1.

Let the irregular block division position parameters of image blocks A, B, C, D and E be posA, posB, posC, posD and posE.

If the image block A, or, B, or, C, or D is not available or does not use irregular block division, then set posA, or, posB, or, posC, or, the value of posD is a sufficiently large number, such as 10000.

The reference motion vector (or predictive motion vector, predicted motion vector) of the current image block E is mvpE.

The calculation method of mvpE is as follows:

If the values of refA and refE are equal and A and E adopt the same irregular division method, set the value of CondionA to 1, otherwise the value of CondionA is 0, the expression is: ConditonA=(refA==refE)&&(ModeA==ModeE).

If the values of refB and refE are equal and B and E adopt the same irregular division method, then set the value of CondionB to 1, otherwise the value of CondionB is 0, the expression is: ConditonB=(refB==refE)&&(ModeB==ModeE).

If the values of refC and refE are equal and C and E adopt the same irregular division method, then set the value of CondionC to 1, otherwise the value of CondionC is 0, the expression is: ConditonC=(refC==refE)&&(ModeC==ModeE).

If the values of refD and refE are equal and D and E adopt the same irregular division method, then set the value of CondionD to 1, otherwise the value of CondionD is 0, the expression is: ConditonD=(refD==refE)&&(ModeD==ModeE).

If ConditonA==1 and ConditonB==0 and ConditonC==0, then mvpE=mvA;

If ConditonA==0 and ConditonB==1 and ConditonC==0, then mvpE=mvB;

If ConditonA==0 and ConditonB==0 and ConditonC==1, then mvpE=mvC;

If ConditonA, ConditonB and ConditonC have more than one value of 1, then get the position parameter posX of the image block X whose ConditonX value is 1 (X can be A, or, B, or, C), calculate posE and these respectively The absolute value of the difference of the image block position parameter (posX), the motion vector mvX (X can be A, or, B, or, C) of the absolute value image block that can obtain the smallest difference is assigned to mvpE;

If ConditonA, ConditonB and ConditonC have more than one value of 1, and the position parameter posX (X can be A, or, B, or, C) of the image block X is equal and the absolute value of the difference with posE can be obtained The absolute value of the minimum difference, then mvpE is equal to the average value of the motion vectors of these image blocks, or mvpE is equal to the average value of the motion vectors of these image blocks and rounded.

Correspondingly, please refer to FIG. 3 , the embodiment of the present invention also provides a corresponding encoding device to implement the encoding method provided by the present invention, which includes:

The irregular division module 210 is configured to select an irregular division mode among a plurality of preset irregular division modes to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, and according to the sub-image The pixel value of the block obtains the reference image block of the sub-image block;

A motion vector obtaining module 220, configured to obtain a motion vector of a sub-image block of the image block to be encoded according to the reference image block;

The reference motion vector obtaining module 230 obtains the reference motion vector of the sub-image block according to the coded surrounding blocks of the sub-image block of the image block to be encoded;

The obtaining the reference motion vector of the sub-image block according to the encoded peripheral blocks of the sub-image block of the image block to be encoded includes:

Select surrounding blocks that meet preset conditions;

Obtaining the reference motion vector of the image block according to the surrounding blocks meeting the preset conditions, wherein,

The preset conditions include:

The surrounding blocks adopt the same irregular division method as the image block to be coded;

The reference image block of the peripheral block and the reference image block of the sub-image block are in the same image.

Then, if there is only one peripheral block that satisfies the preset condition, the reference motion vector of the sub-image block is equal to the motion vector of the peripheral block that satisfies the condition.

If more than one of the coded peripheral blocks meets the above conditions, the reference motion vector of the sub-image block is equal to the one whose irregularly divided position parameter is closest to the image block to be coded among the peripheral blocks that meet the conditions The motion vector of the block, in the embodiment of the present invention, the closest may be the smallest difference between position parameters, or the smallest difference between the absolute values of position parameters.

If the position parameters of multiple satisfying surrounding blocks are the same and are also close to the position parameters of the image block to be encoded, the reference motion vector of the sub-image block is the same as that of the multiple satisfying surrounding blocks. The average value of the motion vectors, or the average value of the motion vectors of the plurality of surrounding blocks satisfying the condition is rounded.

A motion vector difference obtaining module 240, configured to make a difference between the motion vector and the reference motion vector to obtain a motion vector difference;

The motion vector difference value writing module 250 is configured to write the motion vector difference value into the code stream.

Referring to Fig. 4, the present invention also provides a decoding method, which includes

Step 301: Obtain the motion vector difference of the sub-image block of the image block to be decoded from the code stream;

Step 302: Obtain a reference motion vector of the sub-image block according to the decoded surrounding blocks of the sub-image block of the image block to be decoded;

specific:

First, it is judged whether the decoded peripheral block meets the preset condition. In the embodiment of the present invention, the preset condition includes:

The peripheral block adopts the same irregular division method as the image block to be decoded;

The reference image block of the peripheral block and the reference image block of the sub-image block are in the same image.

Then, if there is only one peripheral block that satisfies the preset condition, the reference motion vector of the sub-image block is equal to the motion vector of the peripheral block that satisfies the condition.

If more than one of the decoded peripheral blocks meets the above conditions, the reference motion vector of the sub-image block is equal to the motion of the peripheral block whose position parameter is closest to the image block to be decoded among the peripheral blocks that satisfy the condition Vector, in the embodiment of the present invention, the closest may be that the difference between the position parameters is the smallest, or the absolute value of the difference between the position parameters is the smallest.

If the position parameters of multiple surrounding blocks satisfying the condition are the same, the reference motion vector of the sub-image block is the average value of the motion vectors of the multiple surrounding blocks satisfying the condition, or the average value of the motion vectors of the surrounding blocks Rounding.

Step 303: Obtain the motion vector of the sub-image block according to the reference motion vector of the sub-image block and the motion vector difference;

The motion vector difference obtained according to the motion vector difference and the reference motion vector can be obtained by inverse operation according to the generation method of the motion vector difference at the encoding end, which can be the motion vector difference and the reference motion vector sum, or the product of the motion vector difference and the reference motion vector, etc.

Step 304: Find a reference image block of the sub-image block according to the motion vector, and decode the sub-image block according to the reference sub-image block.

It should be pointed out that the image blocks, sub-image blocks and reference image blocks of surrounding blocks in the above solution are nouns defined for convenience of description. The image block in the above solutions refers to the basic unit of image coding, and one image block may contain one to several sub-image blocks. In the existing video codec technology, the image block can be a macroblock with a size of 16x16. At this time, the sub-image block is called a sub-block, and the size can be 16x16, 16x8, 8x16, 8x8, 8x4, 4x8, 4x4; A block can also be a super macroblock, with a size of 32x32. At this time, the sub-image block is a macroblock or a sub-block in a macroblock, and the size can be 32x32, 32x16, 16x32, 16x16, 16x8, 8x16, 8x8, 8x4, 4x8, 4x4.

The surrounding blocks refer to image blocks or sub-image blocks adjacent to the current image block or the current sub-image block. The positional relationship between it and the current image block, or the current sub-image block can be adjacent, or pixel values at a fixed interval. The surrounding blocks of the current sub-image block may be image blocks or sub-image blocks.

The reference image blocks of the surrounding blocks may be the above-mentioned image blocks, or the above-mentioned sub-image blocks. Therefore, if the surrounding blocks of the current sub-image block are image blocks, the reference image blocks of the surrounding blocks of the current sub-image block can be image blocks; if the surrounding blocks of the current sub-image block are sub-image blocks, the current sub-image The reference image blocks of the peripheral blocks of the block may be sub-image blocks.

further,

Referring to FIG. 3 , image blocks A, B, C, D and E are provided. Where E is the current image block, A is the left adjacent block of E, B is the upper adjacent block of E, C is the upper right adjacent block of E, and D is the upper left adjacent block of E.

Let the motion vectors of image blocks A, B, C, and D be mvA, mvB, mvC, and mvD. Suppose the reference frame index values of image blocks A, B, C, D and E are refA, refB, refC, refD and refE.

The reference frame index value refers to the position in the video sequence of the image where the reference image block of the image block is located. When the reference frame index values of two image blocks are the same, the reference image blocks of the two image blocks are located in the same image.

It is assumed that the irregular block division methods of the image blocks A, B, C, D and E are ModeA, ModeB, ModeC, ModeD and ModeE.

If the image block A, or, B, or, C, or D is unavailable or does not use irregular block division, set the value of ModeA, or, ModeB, or, ModeC, or, ModeD to -1.

Let the irregular block division position parameters of image blocks A, B, C, D and E be posA, posB, posC, posD and posE.

If the image block A, or, B, or, C, or D is not available or does not use irregular block division, then set posA, or, posB, or, posC, or, the value of posD is a sufficiently large number, such as 10000.

The reference motion vector (or predictive motion vector, predicted motion vector) of the current image block E is mvpE.

The calculation method of mvpE is as follows:

If the values of refA and refE are equal and A and E adopt the same irregular division method, set the value of CondionA to 1, otherwise the value of CondionA is 0, the expression is: ConditonA=(refA==refE)&&(ModeA==ModeE).

If the values of refB and refE are equal and B and E adopt the same irregular division method, then set the value of CondionB to 1, otherwise the value of CondionB is 0, the expression is: ConditonB=(refB==refE)&&(ModeB==ModeE).

If the values of refC and refE are equal and C and E adopt the same irregular division method, then set the value of CondionC to 1, otherwise the value of CondionC is 0, the expression is: ConditonC=(refC==refE)&&(ModeC==ModeE).

If the values of refD and refE are equal and D and E adopt the same irregular division method, then set the value of CondionD to 1, otherwise the value of CondionD is 0, the expression is: ConditonD=(refD==refE)&&(ModeD==ModeE).

If ConditonA==1 and ConditonB==0 and ConditonC==0, then mvpE=mvA;

If ConditonA==0 and ConditonB==1 and ConditonC==0, then mvpE=mvB;

If ConditonA==0 and ConditonB==0 and ConditonC==1, then mvpE=mvC;

If ConditonA, ConditonB and ConditonC have more than one value of 1, then get the position parameter posX of the image block X whose ConditonX value is 1 (X can be A, or, B, or, C), calculate posE and these respectively The absolute value of the difference of the image block position parameter (posX), the motion vector mvX (X can be A, or, B, or, C) of the absolute value image block that can obtain the smallest difference is assigned to mvpE;

If ConditonA, ConditonB and ConditonC have more than one value of 1, and the position parameter posX (X can be A, or, B, or, C) of the image block X is equal and the absolute value of the difference with posE can be obtained The absolute value of the minimum difference, then mvpE is equal to the average value of the motion vectors of these image blocks, or mvpE is equal to the average value of the motion vectors of these image blocks and rounded.

Please refer to FIG. 5 , the embodiment of the present invention also provides a decoding device 400 for executing the above decoding program, which includes:

The motion vector difference obtaining module 410 is used to obtain the motion vector difference of the sub-image block of the image block to be decoded from the code stream;

A reference motion vector obtaining module 420, configured to obtain a reference motion vector of a sub-image block of the image block to be decoded according to decoded surrounding blocks of the sub-image block;

specific:

First, the reference motion vector obtaining module 420 judges whether the decoded peripheral blocks meet the preset conditions. In the embodiment of the present invention, the preset conditions include:

The peripheral block adopts the same irregular division method as the image block to be decoded;

The reference image block of the peripheral block and the reference image block of the sub-image block are in the same image.

Then, if there is only one peripheral block that satisfies the preset condition, the reference motion vector of the sub-image block is equal to the motion vector of the peripheral block that satisfies the condition.

If more than one of the decoded surrounding blocks satisfies the above conditions, the reference motion vector of the sub-image block is equal to the surrounding block whose irregularly divided position parameter is closest to the image block to be decoded among the surrounding blocks that meet the conditions The motion vector of the block, in the embodiment of the present invention, the closest may be the smallest difference between position parameters, or the smallest difference between the absolute values of position parameters.

If the position parameters of multiple satisfying surrounding blocks are the same and are also close to the position parameters of the image block to be decoded, the reference motion vector of the sub-image block is the reference motion vector of the multiple satisfying surrounding blocks The average value of the motion vectors, or the average value of the motion vectors of the plurality of surrounding blocks satisfying the condition is rounded.

A motion vector obtaining module 430, configured to obtain the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block;

The decoding module 440 is configured to search for a reference image block of the sub-image block according to the motion vector, and decode the sub-image block according to the reference sub-image block.

In the encoding method or decoding method provided by the embodiment of the present invention, according to the continuity of the image content or texture, the present invention also provides another encoding/decoding method based on the sub-image block of the image block to be encoded or decoded A method for obtaining a reference motion vector of a sub-image block, the method comprising:

First, judging whether the encoded/decoded peripheral blocks meet a preset condition;

In the embodiment of the present invention, the preset condition includes: the peripheral block adopts the same irregular division method as the image block to be encoded/decoded.

Wherein, when the image block to be coded/decoded adopts the horizontal division method of dividing the image block into two sub-image blocks by the horizontal dividing line shown in Fig. 2(d), if the image block to be coded/decoded The left adjacent block also adopts the horizontal partition method, then obtain the reference motion vector of the current image block according to the motion vector of the left adjacent block of the current coded/decoded image block; if the left adjacent block of the image block to be encoded/decoded does not have If the horizontal division mode is adopted and the upper adjacent block of the currently encoded image block adopts the horizontal division mode, then the reference motion vector of the image block to be encoded/decoded is obtained according to the motion vector of the upper adjacent block of the image block to be encoded/decoded.

When the image block to be coded/decoded adopts the vertical division method of dividing the image block into two sub-image blocks by the vertical division line shown in Fig. 2(c), if the upper adjacent block of the image block to be coded/decoded The vertical division method is also adopted, and the reference motion vector of the image block to be encoded/decoded is obtained according to the motion vector of the upper adjacent block of the image block to be encoded/decoded; if the upper adjacent block of the image block to be encoded/decoded If the block does not use the vertical division method and the left adjacent block of the image block to be encoded/decoded adopts the vertical division method, then the image to be encoded/decoded is obtained according to the motion vector of the adjacent image block to the left of the image block to be encoded/decoded The block's reference motion vector.

Further, the method for obtaining the reference motion vector of the current image block according to the motion vectors of the adjacent image blocks above or to the left of the image block to be encoded/decoded may be:

When the image block to be encoded/decoded adopts the horizontal division method of dividing the image block into two sub-image blocks by the horizontal dividing line shown in Figure 2(d), if the left phase of the image block to be encoded/decoded The adjacent block also adopts the horizontal division method, then the reference motion vector of the image block to be encoded/decoded is equal to the motion vector of the left adjacent block of the image block to be encoded/decoded; if the left adjacent block of the image block to be encoded/decoded If the adjacent block does not adopt the horizontal division method and the upper adjacent block of the current image block to be encoded/decoded adopts the horizontal division method, then the reference motion vector of the image block to be encoded/decoded is equal to the reference motion vector of the image block to be encoded/decoded The motion vector of the upper adjacent block;

When the image block to be encoded/decoded is divided into two sub-image blocks by the vertical dividing line shown in Fig. 2(c), if the upper adjacent block of the image block to be encoded/decoded is also If the vertical division method is adopted, the reference motion vector of the image block to be encoded/decoded is equal to the motion vector of the adjacent block on the image block to be encoded/decoded; if the upper adjacent block of the image block to be encoded/decoded is not used Vertical division method and the left adjacent block of the image block to be encoded/decoded adopts the vertical division method, then the reference motion vector of the image block to be encoded/decoded is equal to the motion vector of the left adjacent block of the image block to be encoded/decoded .

It should be understood that the horizontal division method and the vertical division method described in this embodiment include the irregular division method of dividing the image block into two rectangular sub-image blocks of different sizes, and the even division of the image block into two equal sub-image blocks. The regular division of image blocks (16x8 or 8x16).

When the image block to be encoded/decoded is divided into a triangle and a pentagon using the method shown in Figure 2(a) or Figure 2(b), the surrounding area of the image block to be encoded/decoded can be The reference motion vector of the image block to be encoded/decoded is obtained by using the motion vector formula of the surrounding blocks in the same irregular division mode as the image block to be encoded/decoded.

As an optional embodiment, the present invention also provides a method for obtaining the reference motion vector of the current image block according to the motion vectors of the adjacent image blocks above or to the left of the image block to be encoded/decoded:

According to the reference frame index value of the current image block, and the reference frame index value of the adjacent block on the current image block and its motion vector, obtain the reference motion amount of the current image block; or

The reference motion vector of the current image block is obtained according to the reference frame index value of the current image block, and the reference frame index value of the left adjacent block of the current image block (ie, the peripheral block on the left side of the current image block) and its motion vector.

A further implementation will be given below.

In the following implementation schemes, since the surrounding blocks of the current image block or sub-image block can be image blocks or sub-image blocks, they are collectively referred to as image blocks for the convenience of description. The sizes of all image blocks may be inconsistent, and the shapes or division methods of all image blocks It can be inconsistent, and the size of the image block is determined according to the method in the prior art.

As shown in Figure 3, there are image blocks A, B, C, D and E. Where E is the current image block, A is the left adjacent block of E, B is the upper adjacent block of E, C is the upper right adjacent block of E, and D is the upper left adjacent block of E.

The current image block adopts irregular division, and the image block is divided into two sub-image blocks. The reference motion vector (or predictive motion vector, predicted motion vector) of the first or second sub-image block E of the current image block is mvpE.

If the image block A, or, B, or, C, or D is unavailable or does not use irregular block division, set the value of ModeA, or, ModeB, or, ModeC, or, ModeD to -1.

If image blocks A, B, C, and D are divided irregularly, then image blocks A, B, C, and D are divided into two sub-image blocks. , the motion vectors of D are mvA, mvB, mvC, mvD, and the reference frame index values of the sub-image blocks corresponding to the above-mentioned image blocks A, B, C, D, and E are refA, refB, refC, refD, and refE.

The reference frame index value refers to the position in the video sequence of the image where the reference image block of the image block or sub-image block is located. When the reference frame index values of two image blocks or sub-image blocks are the same, the reference image blocks of the two image blocks or sub-image blocks are located in the same image.

If ModeE is a horizontal division method, and ModeA is also a horizontal division method, the calculation method of mvpE can be as follows:

mvpE=mvA

Or, mvpE=((refE+1)×mvA×768/(refA+1)+384)/768

The above publicity ((refE+1)×mvA×768/(refA+1)+384)/768 actually scales mvA, and scales the motion vector of block A to the reference image pointed to by the current block E.

If ModeE is a horizontal division method, ModeA is not a horizontal division method, and ModeB is a horizontal division method, then the method for obtaining mvpE can be:

mvpE=mvB

Or, mvpE=((refE+1)×mvB×768/(refB+1)+384)/768

The above formula ((refE+1)×mvB×768/(refB+1)+384)/768 actually scales mvB, and scales the motion vector of block B to the reference image pointed to by the current block E.

If ModeE is a vertical division method, and ModeB is also a vertical division method, the calculation method of mvpE can be as follows:

mvpE=mvB

Or, mvpE=((refE+1)×mvB×768/(refB+1)+384)/768

The above publicity ((refE+1)×mvB×768/(refB+1)+384)/768 actually scales mvB, and scales the motion vector of block B to the reference image pointed to by the current block E.

If ModeE is a vertical division mode, ModeB is not a vertical division mode, and ModeA is a vertical division mode, then the mvpE calculation method can be as follows:

mvpE=mvA

Or, mvpE=((refE+1)×mvA×768/(refA+1)+384)/768

The above formula ((refE+1)×mvA×768/(refA+1)+384)/768 actually scales mvA, and scales the motion vector of block A to the reference image pointed to by the current block E.

The present invention also provides other optional embodiments for obtaining the reference vector of the image block. In the following implementation schemes, since the surrounding blocks of the current image block or sub-image block can be image blocks or sub-image blocks, for convenience The descriptions are collectively referred to as image blocks. The sizes of all image blocks may be inconsistent, and the shapes or division methods of all image blocks may be inconsistent. The size of image blocks is determined according to methods in the prior art.

As shown in Figure 3, there are image blocks A, B, C, D and E. Where E is the current image block, A is the left adjacent block of E, B is the upper adjacent block of E, C is the upper right adjacent block of E, and D is the upper left adjacent block of E.

The current image block adopts irregular division, and the image block is divided into two sub-image blocks. The reference motion vector (or predictive motion vector, predicted motion vector) of the first or second sub-image block E of the current image block is mvpE.

If the image block A, or, B, or, C, or D is unavailable or does not use irregular block division, set the value of ModeA, or, ModeB, or, ModeC, or, ModeD to -1.

If the image blocks A, B, C, and D have adopted an irregular division method, then the image blocks A, B, C, and D are divided into two sub-image blocks (, assuming that the image blocks A, B, and The motion vectors of C and D are mvA, mvB, mvC, and mvD, and the reference frame index values of the sub-image blocks corresponding to the above-mentioned image blocks A, B, C, D, and E are refA, refB, refC, refD, and refE).

If ModeE is a horizontal division mode, and ModeA is also a horizontal division mode, then the reference motion vectors of the first and second sub-image blocks of image block E are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), and image block A The reference motion vectors of the first and second sub-image blocks are (mvA_y1, mvA_x1) and (mvA_y2, mvA_x2), (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2) can be obtained by:

mvpE_y1=mvA_y1

mvpE_x1=mvA_x1

mvpE_y2=mvA_y2

mvpE_x2 = mvA_x2

If ModeE is the horizontal division mode, ModeA is not the horizontal division mode, and ModeB is the horizontal division mode. The reference motion vectors of the first and second sub-image blocks of image block E are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), and the reference motion vectors of the first and second sub-image blocks of image block B are (mvB_y1 , mvB_x1) and (mvB_y2, mvB_x2). At this time, the lower half of the image block B (the upper adjacent block) (that is, the second sub-image block of the image block B) and the current image block E are more likely to belong to the same object, and the motion vectors of the two image blocks are correlated higher degree. Therefore, the calculation method of (mvpE_y1, mvpE_x1) can be:

mvpE_y1=mvB_y2

mvpE_x1=mvB_x2

In the above scenario, the motion vector of the second sub-image block of the image block E (ie, the lower half of the image block E) is less correlated with the motion vector of the image block B. At this time, the motion vector of the left block of the sub-image block is used, That is, the motion vector of the part of the image block A adjacent to the image block E is more appropriate. Here, let the motion vector of the part of the image block A adjacent to the second sub-image block of the image block E be (mvA_yclose, mvA_xclose). Therefore, the calculation method of (mvpE_y2, mvpE_x2) can be:

mvpE_y2 = mvA_yclose

mvpE_x2 = mvA_xclose

If ModeE is a vertical division mode, and ModeA is also a vertical division mode, the reference motion vectors of the first and second sub-image blocks of image block E are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), and image block B The reference motion vectors of the first and second sub-image blocks are (mvB_y1, mvB_x1) and (mvB_y2, mvB_x2), (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2) can be obtained as follows:

mvpE_y1=mvB_y1

mvpE_x1=mvB_x1

mvpE_y2=mvB_y2

mvpE_x2=mvB_x2

If ModeE adopts the vertical division mode, ModeB neither adopts the vertical division mode, but ModeA adopts the vertical division mode. The reference motion vectors of the first and second sub-image blocks of image block E are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), and the reference motion vectors of the first and second sub-image blocks of image block B are (mvA_y1 , mvA_x1) and (mvA_y2, mvA_x2). At this time, the right half of image block A (the left adjacent block) (that is, the second sub-image block of image block A) and the current image block E are more likely to belong to the same object, and the motion vectors of these two image blocks are associated higher degree. Therefore, the calculation method of (mvpE_y1, mvpE_x1) can be:

mvpE_y1=mvA_y2

mvpE_x1=mvA_x2

In the above scenario, the motion vector of the second sub-image block of image block E (that is, the right part of image block E) is less correlated with the motion vector of image block A, and at this time, the motion vector of the adjacent block on the sub-image block is used , that is, the motion vector of the part of the upper adjacent block B that is close to the image block E is more appropriate. Let the motion vector of the part of the image block B close to the image block E be (mvB_yclose, mvB_xclose), therefore, the method of obtaining (mvpE_y2, mvpE_x2) can be:

mvpE_y2 = mvA_yclose

mvpE_x2 = mvA_xclose

If ModeE is the upper left to lower right division mode, the division method shown in Figure 2(a) is adopted, and image blocks A (left adjacent block), B (upper adjacent block), and D (upper left adjacent block) have And there is only one that adopts the same division mode as ModeE, then the motion vectors of the first and second image blocks of the image block that adopts the same division mode as ModeE are respectively (mvX_y1, mvX_x1), (mvX_y2, mvX_x2) , the reference motion vectors of the first and second sub-image blocks of the image block E are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), then (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2) are calculated as follows:

mvpE_y1=mvX_y1

mvpE_x1=mvX_x1

mvpE_y2=mvX_y2

mvpE_x2=mvX_x2

If more than one image block in A, B, and D adopts the same division mode as ModeE, but the reference frame index value of only one image block is the same as the reference frame index value of the current image block, then the division of the image block with the current image block is set The motion vectors of the first and second sub-image blocks of the image block with the same method and the same reference frame index value are (mvX_y1, mvX_x1), (mvX_y2, mvX_x2), and the reference of the first and second sub-image blocks of the image block E The motion vectors are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), X can be A or B or D, then (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2) are calculated as:

mvpE_y1=mvX_y1

mvpE_x1=mvX_x1

mvpE_y2=mvX_y2

mvpE_x2=mvX_x2

If ModeE is the upper right to lower left division mode, that is, the division method shown in Figure 2(b) is adopted, and one and only one of the image blocks A, B, and C adopts the same division mode as ModeE, then it is assumed that the mode is the same as ModeE The block motion vectors of the first and second sub-images of the image block of the division mode are (mvX_y1, mvX_x1), (mvX_y2, mvX_x2), and the reference motion vectors of the first and second sub-image blocks of the image block E are ( mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), x can be A, B or C, then (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2) are calculated as:

mvpE_y1=mvX_y1

mvpE_x1=mvX_x1

mvpE_y2=mvX_y2

mvpE_x2=mvX_x2

If there is more than one image block in A, B, and C that adopts the same division mode as ModeE, but the reference frame index value of only one image block is the same as the reference frame index value of the current image block, then the division of the current image block is set The motion vectors of the first and second sub-image blocks of the image block with the same mode and the same reference frame index value are (mvX_y1, mvX_x1), (mvX_y2, mvX_x2), X can be A or B or C, and the image block E is the first The reference motion vectors of one and the second sub-image block are (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2), then (mvpE_y1, mvpE_x1) and (mvpE_y2, mvpE_x2) are calculated as follows:

mvpE_y1=mvX_y1

mvpE_x1=mvX_x1

mvpE_y2=mvX_y2

mvpE_x2=mvX_x2

In the above method, the reference motion vector of the sub-image block can also be obtained according to the encoded/decoded surrounding blocks of the sub-image block of the image block to be encoded/decoded, and the irregular division method of the image block to be encoded/decoded .

The specific method is as follows: the coded/decoded surrounding blocks of the sub-image block of the image block to be encoded/decoded are A, B, C and D shown in FIG. 3 . If the image block to be encoded/decoded adopts a horizontal division method or a vertical division method, the motion vectors of the surrounding blocks A and B are selected to obtain the reference motion vector of the sub-image block of the image block to be encoded/decoded; if the image block to be encoded/decoded The image block to be decoded adopts the division method shown in Fig. 2 (a), then the motion vectors of the peripheral blocks A, B and C are selected for obtaining the reference motion vector of the sub-image block of the image block to be encoded/decoded; The image block to be decoded adopts the division method shown in FIG. 2(b), then the motion vectors of the surrounding blocks A, B and D are selected to obtain the reference motion vectors of the sub-image blocks of the image block to be encoded/decoded.

The encoding and decoding method and device provided by the embodiment of the present invention can simplify the acquisition and calculation of the reference motion vector under certain conditions, save the resources of the encoding and decoding system, and improve the efficiency.

The multi-level switching network and system provided by the embodiments of the present invention are described above in detail. In this paper, specific examples are used to illustrate the principles and implementation modes of the present invention. The descriptions of the above embodiments are only used to help understand the present invention. method and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation and application scope. Invention Limitations.

Claims (28)

1. A coding method, characterized in that, comprising: Select one of the preset irregular division methods to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, and obtain the sub-image blocks according to the pixel values of the sub-image blocks. the reference image block of the image block; According to the reference image block, the motion vector of the sub-image block of the image block to be encoded is obtained; according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded , to obtain the reference motion vector of the sub-image block; Obtaining a motion vector difference according to the motion vector and the reference motion vector; Write the motion vector difference into the code stream, Wherein, according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded, obtaining the reference motion vector of the sub-image block includes: Select the surrounding blocks that meet the preset conditions, and obtain the reference motion vector of the image block according to the surrounding blocks that meet the preset conditions, wherein the preset conditions include: The surrounding block adopts the same irregular division method as the image block to be coded; The reference image block of the peripheral block and the reference image block of the sub-image block are in the same image. the 2. The encoding method according to claim 1, wherein if there is only one peripheral block that meets the preset condition, the reference motion vector of the sub-image block is equal to the motion of the peripheral block that satisfies the condition vector. the 3. The encoding method as claimed in claim 1, wherein if a plurality of the encoded peripheral blocks satisfy the preset condition, the reference motion vector of the sub-image block is equal to The motion vector of the neighboring block with the smallest difference between the position parameter of the surrounding block and the position parameter of the image block to be encoded, or the reference motion vector of the sub-image block is equal to the position parameter and The motion vector of the neighboring block with the smallest absolute value of the position parameter difference of the image block to be coded. the 4. encoding method as claimed in claim 3, it is characterized in that, if the position parameter of a plurality of surrounding blocks that satisfy the condition is identical and the position parameter in the surrounding block that satisfies the condition is the same as the position parameter of the image block to be encoded The absolute value of the difference is the smallest, or if there are multiple surrounding blocks that meet the condition and the position parameters are the same and the absolute value of the difference between the position parameters in the surrounding blocks that meets the condition and the position parameter of the image block to be encoded is the smallest, then the sub The reference motion vector of the image block is obtained by taking the average value of the motion vectors of the plurality of surrounding blocks satisfying the conditions, or the reference motion vector of the sub-image block is obtained by taking the average value of the motion vectors of the plurality of surrounding blocks satisfying the conditions all. the 5. A coding method, characterized in that, comprising: Select one of the preset irregular division methods to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, and obtain the sub-image blocks according to the pixel values of the sub-image blocks. the reference image block of the image block; According to the reference image block, the motion vector of the sub-image block of the image block to be encoded is obtained; according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded , to obtain the reference motion vector of the sub-image block; Obtaining a motion vector difference according to the motion vector and the reference motion vector; Write the motion vector difference into the code stream, Wherein, according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded, obtaining the reference motion vector of the sub-image block includes: When the irregular division method of the image block to be encoded is a horizontal division method, If the irregular division mode of the left adjacent block of the image block to be encoded is a horizontal division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the left adjacent block; If the irregular division mode of the left adjacent block of the image block to be encoded is not a horizontal division mode, and the irregular division mode of the upper adjacent block of the image block to be encoded is a horizontal division mode, then according to The reference motion vector of the sub-image block of the current image block is obtained from the motion vector of the sub-image block of the upper adjacent block. the 6. A coding method, characterized in that, comprising: Select one of the preset irregular division methods to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, and obtain the sub-image blocks according to the pixel values of the sub-image blocks. the reference image block of the image block; According to the reference image block, the motion vector of the sub-image block of the image block to be encoded is obtained; according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded , to obtain the reference motion vector of the sub-image block; Obtaining a motion vector difference according to the motion vector and the reference motion vector; Write the motion vector difference into the code stream, Wherein, according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded, obtaining the reference motion vector of the sub-image block includes: When the irregular division method of the image block to be encoded is a vertical division method, If the irregular division mode of the upper adjacent block of the image block to be encoded is a vertical division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the upper adjacent block; If the irregular division mode of the upper adjacent block of the image block to be encoded is not vertical division mode, and the irregular division mode of the left adjacent block of the image block to be encoded is vertical division mode, according to the left The motion vector of the sub-image block of the adjacent block obtains the reference motion vector of the sub-image block of the current image block. the 7. A coding method, characterized in that, comprising: Select one of the preset irregular division methods to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, and obtain the sub-image blocks according to the pixel values of the sub-image blocks. the reference image block of the image block; According to the reference image block, the motion vector of the sub-image block of the image block to be encoded is obtained; according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded , to obtain the reference motion vector of the sub-image block; Obtaining a motion vector difference according to the motion vector and the reference motion vector; Write the motion vector difference into the code stream, Wherein, according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division method of the image block to be encoded, obtaining the reference motion vector of the sub-image block includes: When the irregular division method of the image block to be encoded divides the image block into two triangles, or a triangle and a pentagon, or two quadrilaterals, if there is only one peripheral block of the image block to be encoded The image block division method is the same as the irregular division method of the image block to be encoded, then the reference motion vector of the sub-image block of the current image block is obtained according to the motion vector of the sub-image block of the image block; or If there are a plurality of image blocks in the surrounding blocks of the image block to be encoded that are divided in the same manner as the irregular division of the image block to be encoded, then according to the image blocks among the image blocks to be encoded The reference motion vector of the image block to be coded is obtained from the motion vector of the sub-image block of the image block whose reference frame index value is the same. the 8. The encoding method according to claim 5, 6 or 7, wherein the sub-image of the image block to be encoded is obtained according to the motion vector of the sub-image block of the left adjacent block or the upper adjacent block Methods for referencing motion vectors for blocks include: The reference motion vector of the sub-image block of the image block to be encoded is equal to the motion vector of the sub-image block of the left or upper image block; or Scaling the motion vector of the sub-image block of the left adjacent block or the upper adjacent block to the reference image pointed to by the sub-image block of the image block to be encoded, and obtaining the image to be encoded according to the scaled motion vector The reference motion vector of the sub-image block of the image block. the 9. An encoding device, characterized in that, comprising: The irregular division module is used to select an irregular division mode among a plurality of preset irregular division modes to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, according to the sub-image blocks Obtain the reference image block of the sub-image block with the pixel value; A motion vector obtaining module, configured to obtain a motion vector of a sub-image block of the image block to be encoded according to the reference image block; The reference motion vector obtaining module is used for encoding according to the sub-image block of the image block to be encoded According to the irregular division method of the surrounding blocks of the code and the image block to be encoded, the reference motion vector of the sub-image block is obtained, and the obtained sub-image block is obtained according to the coded surrounding blocks of the sub-image block of the image block to be encoded. The reference motion vector includes: selecting a surrounding block that meets the preset condition; and obtaining the reference motion vector of the image block according to the surrounding block that meets the preset condition, wherein the preset condition includes: the surrounding block adopts the same The image blocks to be encoded are divided in the same irregular manner; the reference image blocks of the surrounding blocks and the reference image blocks of the sub-image blocks are in the same image; A motion vector difference obtaining module, configured to make a difference between the motion vector and the reference motion vector to obtain a motion vector difference; The motion vector difference value writing module is used to write the motion vector difference value into the code stream. the 10. The encoding device according to claim 9, wherein if there is only one peripheral block that satisfies a preset condition, the reference motion vector of the sub-image block is equal to the motion of the peripheral block that satisfies the condition vector. the 11. The encoding device according to claim 9, wherein if more than one of the encoded peripheral blocks satisfy the preset condition, the reference motion vector of the sub-image block is equal to The motion vector of the neighboring block with the smallest difference between the irregularly divided position parameter and the position parameter of the image block to be coded, or the position parameter in the surrounding block that satisfies the condition and the position parameter of the image block to be coded The motion vector of the surrounding block whose absolute value of the difference is the smallest. the 12. The encoding device as claimed in claim 11, wherein if the position parameters of a plurality of surrounding blocks satisfying the condition are the same and the position parameters in the surrounding blocks satisfying the condition are the same as the position parameters of the image block to be encoded The absolute value of the difference is the smallest, or if there are multiple surrounding blocks that meet the condition and the position parameters are the same and the absolute value of the difference between the position parameters in the surrounding blocks that meets the condition and the position parameter of the image block to be encoded is the smallest, then the sub The reference motion vector of the image block takes the average value of the motion vectors of the plurality of surrounding blocks that meet the conditions, or the reference motion vector of the sub-image block takes the average value of the motion vectors of the multiple surrounding blocks that meet the conditions Rounding. the 13. An encoding device, characterized in that, comprising: The irregular division module is used to select an irregular division mode among a plurality of preset irregular division modes to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, according to the sub-image blocks Obtain the reference image block of the sub-image block with the pixel value; A motion vector obtaining module, configured to obtain a motion vector of a sub-image block of the image block to be encoded according to the reference image block; The reference motion vector obtaining module is used to obtain the reference motion vector of the sub-image block according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division mode of the image block to be encoded, Wherein, said obtaining the reference motion vector of the sub-image block according to the encoded peripheral blocks of the sub-image block of the image block to be encoded comprises: When the irregular division method of the image block to be encoded is a horizontal division method, If the irregular division mode of the left adjacent block of the image block to be encoded is a horizontal division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the left adjacent block; If the irregular division mode of the left adjacent block of the image block to be encoded is not a horizontal division mode, and the irregular division mode of the upper adjacent block of the image block to be encoded is a horizontal division mode, then according to The reference motion vector of the sub-image block of the current image block is obtained from the motion vector of the sub-image block of the upper adjacent block. the 14. An encoding device, characterized in that, comprising: The irregular division module is used to select an irregular division mode among a plurality of preset irregular division modes to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, according to the sub-image blocks Obtain the reference image block of the sub-image block with the pixel value; A motion vector obtaining module, configured to obtain a motion vector of a sub-image block of the image block to be encoded according to the reference image block; The reference motion vector obtaining module is used to obtain the reference motion vector of the sub-image block according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division mode of the image block to be encoded, Wherein, said obtaining the reference motion vector of the sub-image block according to the encoded peripheral blocks of the sub-image block of the image block to be encoded comprises: When the irregular division method of the image block to be encoded is a vertical division method, If the irregular division mode of the upper adjacent block of the image block to be encoded is a vertical division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the upper adjacent block; If the irregular division mode of the upper adjacent block of the image block to be encoded is not vertical division mode, and the irregular division mode of the left adjacent block of the image block to be encoded is vertical division mode, according to the left The motion vector of the sub-image block of the adjacent block obtains the reference motion vector of the sub-image block of the current image block. the 15. An encoding device, characterized in that, comprising: The irregular division module is used to select an irregular division mode among a plurality of preset irregular division modes to irregularly divide the image block to be coded to obtain at least two different sub-image blocks, according to the sub-image blocks Obtain the reference image block of the sub-image block with the pixel value; A motion vector obtaining module, configured to obtain a motion vector of a sub-image block of the image block to be encoded according to the reference image block; The reference motion vector obtaining module is used to obtain the reference motion vector of the sub-image block according to the coded peripheral blocks of the sub-image block of the image block to be encoded and the irregular division mode of the image block to be encoded, Wherein, said obtaining the reference motion vector of the sub-image block according to the encoded peripheral blocks of the sub-image block of the image block to be encoded comprises: When the irregular division method of the image block to be encoded divides the image block into two triangles, or a triangle and a pentagon, or two quadrilaterals, if there is only one peripheral block of the image block to be encoded The image block division method is the same as the irregular division method of the image block to be encoded, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the image block; or If there are a plurality of image blocks in the surrounding blocks of the image block to be encoded that are divided in the same manner as the irregular division of the image block to be encoded, then according to the image blocks among the image blocks to be encoded The reference motion vector of the image block to be coded is obtained from the motion vector of the sub-image block of the image block whose reference frame index value is the same. the 16. The encoding device according to claim 13, 14 or 15, wherein said obtaining the reference motion vector of the sub-image block according to the coded surrounding blocks of the sub-image block of the image block to be encoded comprises : When the irregular division method of the image block to be encoded divides the image block into two triangles, or a triangle and a pentagon, or two quadrilaterals, if there is only one peripheral block of the image block to be encoded The image block division method is the same as the irregular division method of the image block to be encoded, then the reference motion vector of the sub-image block of the current image block is obtained according to the motion vector of the sub-image block of the image block; or If there are a plurality of image blocks in the surrounding blocks of the image block to be encoded that are divided in the same manner as the irregular division of the image block to be encoded, then according to the image blocks among the image blocks to be encoded The reference motion vector of the image block to be coded is obtained from the motion vector of the sub-image block of the image block whose reference frame index value is the same. the 17. A decoding method, characterized in that, comprising: Obtain the irregular division method of the image block to be decoded from the code stream; Obtain the motion vector difference value of the sub-image block of the image block to be decoded from the code stream; Obtain the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block; Obtaining the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; obtaining a reference image block of the sub-image block according to the motion vector, and decoding the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block includes: Select a surrounding block that meets the preset condition; obtain a reference motion vector of the image block according to the surrounding block that meets the preset condition, wherein, The preset conditions include: the peripheral block adopts the same irregular division method as the image block to be decoded; the reference image block of the peripheral block and the reference image block of the sub-image block are in the same image . the 18. The decoding method according to claim 17, wherein if there is only one peripheral block that satisfies the preset condition, the reference motion vector of the sub-image block is equal to the motion of the peripheral block that satisfies the condition vector. the 19. The decoding method according to claim 17, wherein if multiple of the decoded peripheral blocks meet the preset condition, the reference motion vector of the sub-image block is equal to The motion vector of the neighboring block with the smallest difference between the position parameters in the surrounding blocks and the position parameters of the image block to be decoded, or the difference between the position parameters in the surrounding blocks that meet the conditions and the position parameters of the image block to be decoded The motion vector of the surrounding block with the smallest absolute value. the 20. The decoding method as claimed in claim 19, wherein if the position parameters of a plurality of surrounding blocks satisfying the conditions are the same and the position parameters in the surrounding blocks satisfying the conditions are the same as the position parameters of the image block to be decoded The absolute value of the difference is the smallest, or if there are multiple surrounding blocks that meet the condition with the same position parameters and the absolute value of the difference between the position parameters in the surrounding blocks that meet the conditions and the position parameter of the image block to be decoded is the smallest, then the sub The reference motion vector of the image block takes the average value of the motion vectors of the plurality of surrounding blocks that meet the conditions, or the reference motion vector of the sub-image block takes the average value of the motion vectors of the multiple surrounding blocks that meet the conditions Rounding. the 21. A decoding method, characterized in that, comprising: Obtain the irregular division method of the image block to be decoded from the code stream; Obtain the motion vector difference value of the sub-image block of the image block to be decoded from the code stream; Obtain the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block; Obtaining the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; obtaining a reference image block of the sub-image block according to the motion vector, and decoding the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block includes: When the irregular division method of the image block to be decoded is a horizontal division method, If the irregular division mode of the left adjacent block of the image block to be decoded is a horizontal division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the left adjacent block; If the irregular division mode of the left adjacent block of the image block to be decoded is not a horizontal division mode, and the irregular division mode of the upper adjacent block of the image block to be decoded is a horizontal division mode, then according to The reference motion vector of the sub-image block of the current image block is obtained from the motion vector of the sub-image block of the upper adjacent block. the 22. A decoding method, characterized in that, comprising: Obtain the irregular division method of the image block to be decoded from the code stream; Obtain the motion vector difference value of the sub-image block of the image block to be decoded from the code stream; Obtain the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block; Obtaining the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; obtaining a reference image block of the sub-image block according to the motion vector, and decoding the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block includes: When the irregular division mode of the image block to be decoded is vertical division mode, If the irregular division mode of the upper adjacent block of the image block to be decoded is a vertical division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the upper adjacent block; If the irregular division mode of the upper adjacent block of the image block to be decoded is not a vertical division mode, and the irregular division mode of the left adjacent block of the image block to be decoded is a vertical division mode, according to the left The motion vector of the sub-image block of the adjacent block obtains the reference motion vector of the sub-image block of the current image block. the 23. A decoding method, characterized in that, comprising: Obtain the irregular division method of the image block to be decoded from the code stream; Obtain the motion vector difference value of the sub-image block of the image block to be decoded from the code stream; Obtain the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block; Obtaining the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; obtaining a reference image block of the sub-image block according to the motion vector, and decoding the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block from the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block includes: When the irregular division method of the image block to be decoded divides the image block into two triangles, or a triangle and a pentagon, or two quadrilaterals, if there is only one peripheral block of the image block to be decoded The image block division method is the same as the irregular division method of the image block to be decoded, then the reference motion vector of the sub-image block of the current image block is obtained according to the motion vector of the sub-image block of the image block; or If there are a plurality of image blocks in the peripheral blocks of the image block to be decoded that are divided in the same manner as the irregular division method of the image block to be decoded, then according to the image block among the image blocks to be decoded The reference motion vector of the image block to be decoded is obtained by using the motion vector of the sub-image block of the image block with the same reference frame index value. the 24. The decoding method according to claim 21, 22 or 23, wherein the reference motion of the sub-image block of the image block to be decoded is obtained according to the motion vector of the sub-image block of the left adjacent block or the upper adjacent block Vector methods include: The reference motion vector of the sub-image block of the image block to be decoded is equal to the motion vector of the sub-image block of the left or upper image block; or Scaling the motion vector of the word image block of the left adjacent block or the upper adjacent block to the reference image pointed to by the sub-image block of the image block to be decoded, and obtaining the image block to be decoded according to the scaled motion vector The reference motion vector of the sub-image block of the image block. the 25. A decoding device, characterized in that, comprising: The motion vector difference obtaining module is used to obtain the motion vector difference of the sub-image block of the image block to be decoded from the code stream; The reference motion vector obtaining module is used to obtain the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block to be decoded; A motion vector obtaining module, configured to obtain the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; A decoding module, configured to search for a reference image block of the sub-image block according to the motion vector, and decode the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded includes: When the irregular division method of the image block to be decoded is a horizontal division method, If the irregular division mode of the left adjacent block of the image block to be decoded is a horizontal division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the left adjacent block; If the irregular division mode of the left adjacent block of the image block to be decoded is not a horizontal division mode, and the irregular division mode of the upper adjacent block of the image block to be decoded is a horizontal division mode, then according to The reference motion vector of the sub-image block of the current image block is obtained from the motion vector of the sub-image block of the upper adjacent block. the 26. A decoding device, characterized in that, comprising: The motion vector difference obtaining module is used to obtain the motion vector difference of the sub-image block of the image block to be decoded from the code stream; The reference motion vector obtaining module is used to obtain the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block to be decoded; A motion vector obtaining module, configured to obtain the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; A decoding module, configured to search for a reference image block of the sub-image block according to the motion vector, and decode the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded includes: When the irregular division mode of the image block to be decoded is vertical division mode, If the irregular division mode of the upper adjacent block of the image block to be decoded is a vertical division mode, then obtain the reference motion vector of the sub-image block of the current image block according to the motion vector of the sub-image block of the upper adjacent block; If the irregular division mode of the upper adjacent block of the image block to be decoded is not a vertical division mode, and the irregular division mode of the left adjacent block of the image block to be decoded is a vertical division mode, according to the left The motion vector of the sub-image block of the adjacent block obtains the reference motion vector of the sub-image block of the current image block. the 27. A decoding device, characterized in that, comprising: The motion vector difference obtaining module is used to obtain the motion vector difference of the sub-image block of the image block to be decoded from the code stream; The reference motion vector obtaining module is used to obtain the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded and the irregular division method of the image block to be decoded; A motion vector obtaining module, configured to obtain the motion vector of the sub-image block according to the reference motion vector and the motion vector difference of the sub-image block; A decoding module, configured to search for a reference image block of the sub-image block according to the motion vector, and decode the sub-image block according to the reference sub-image block, Wherein, said obtaining the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded includes: When the irregular division method of the image block to be decoded divides the image block into two triangles, or a triangle and a pentagon, or two quadrilaterals, if there is only one peripheral block of the image block to be decoded The image block division method is the same as the irregular division method of the image block to be decoded, then the reference motion vector of the sub-image block of the current image block is obtained according to the motion vector of the sub-image block of the image block; or If there are a plurality of image blocks in the peripheral blocks of the image block to be decoded that are divided in the same manner as the irregular division method of the image block to be decoded, then according to the image block among the image blocks to be decoded The reference motion vector of the image block to be decoded is obtained by using the motion vector of the sub-image block of the image block with the same reference frame index value. the 28. The decoding device as claimed in claim 25, 26 or 27, wherein said obtaining the reference motion vector of the sub-image block according to the decoded peripheral blocks of the sub-image block of the image block to be decoded comprises : When the irregular division method of the image block to be decoded divides the image block into two triangles, or a triangle and a pentagon, or two quadrilaterals, if there is only one peripheral block of the image block to be decoded The image block division method is the same as the irregular division method of the image block to be decoded, then the reference motion vector of the sub-image block of the current image block is obtained according to the motion vector of the sub-image block of the image block; or If there are a plurality of image blocks in the peripheral blocks of the image block to be decoded that are divided in the same manner as the irregular division method of the image block to be decoded, then according to the image block among the image blocks to be decoded The reference motion vector of the image block to be decoded is obtained by using the motion vector of the sub-image block of the image block with the same reference frame index value. the

Images