Decoding method, encoding method, device and equipment

文档序号：53010 发布日期：2021-09-28 浏览：33次中文

阅读说明：本技术 解码方法、编码方法、装置及设备 (Decoding method, encoding method, device and equipment ) 是由方树清于 2020-03-26 设计创作，主要内容包括：本申请提供一种解码方法、编码方法、装置及设备,该解码方法包括：获取当前块的码流,并从当前块的码流中解析增强型时域运动矢量预测模式的索引信息,基于当前块的第一周围块,确定当前块的匹配块；基于匹配块以及对匹配块进行偏移得到的新的匹配块,确定候选增强型时域运动矢量预测模式,并基于候选增强型时域运动矢量预测模式构建第二时域候选模式列表；基于索引信息,从第二时域候选模式列表中确定增强型时域运动矢量预测模式；基于增强型时域运动矢量预测模式,确定当前块内各子块的运动信息,并基于当前块内各子块的运动信息,对当前块内各子块进行运动补偿。该方法可以提高解码性能。(The application provides a decoding method, an encoding method, a device and equipment, wherein the decoding method comprises the following steps: acquiring a code stream of a current block, analyzing index information of an enhanced time domain motion vector prediction mode from the code stream of the current block, and determining a matching block of the current block based on a first peripheral block of the current block; determining a candidate enhanced temporal motion vector prediction mode based on the matching block and a new matching block obtained by shifting the matching block, and constructing a second temporal candidate mode list based on the candidate enhanced temporal motion vector prediction mode; determining an enhanced temporal motion vector prediction mode from the second temporal candidate mode list based on the index information; and determining the motion information of each subblock in the current block based on the enhanced time domain motion vector prediction mode, and performing motion compensation on each subblock in the current block based on the motion information of each subblock in the current block. The method can improve decoding performance.)

1. A method of decoding, comprising:

acquiring a code stream of a current block, and analyzing index information of an enhanced temporal motion vector prediction mode from the code stream of the current block, wherein the index information is used for identifying the position of the enhanced temporal motion vector prediction mode in a first temporal candidate mode list constructed by a coding end device;

determining a matching block of the current block based on a first surrounding block of the current block;

determining a candidate enhanced temporal motion vector prediction mode based on the matching block and a new matching block obtained by shifting the matching block, and constructing a second temporal candidate mode list based on the candidate enhanced temporal motion vector prediction mode;

determining an enhanced temporal motion vector prediction mode from the second temporal candidate mode list based on the index information;

and determining the motion information of each subblock in the current block based on the enhanced time domain motion vector prediction mode, and performing motion compensation on each subblock in the current block based on the motion information of each subblock in the current block.

2. The method of claim 1, wherein determining the matching block for the current block based on a first surrounding block of the current block comprises:

determining motion information of a first stage based on the first peripheral block;

determining a matching block for the current block based on the motion information of the first stage.

3. The method of claim 2, wherein determining the motion information of the first phase based on the first peripheral block comprises:

determining motion information of a first stage based on the forward motion information and/or the backward motion information of the first peripheral block.

4. The method of claim 3, wherein determining the motion information of the first stage based on the forward motion information and/or the backward motion information of the first peripheral block comprises:

determining the motion information of the first stage as backward motion information of the first surrounding block if the backward motion information of the first surrounding block is available and points to a first frame in a List 1;

if the backward motion information of the first peripheral block is available but the backward motion information of the first peripheral block does not point to the first frame in List1, scaling the backward motion information of the first peripheral block to point to the first frame in List1, and determining that the motion vector of the first stage is a scaled motion vector, and the reference frame index is an index of the first frame in List 1;

If the backward motion information of the first peripheral block is not available but the forward motion information is available, scaling the forward motion information of the first peripheral block to point to a first frame in List1, and determining that the motion vector of the first stage is a scaled motion vector, and the reference frame index is an index of the first frame in List 1;

if neither the forward motion information nor the backward motion information of the first surrounding block is available, determining that the motion vector of the first stage is 0, and the reference frame index is the index of the first frame in List 1;

the reference direction of the motion information in the first stage is the List1 direction.

5. The method of claim 3, wherein determining the motion information of the first stage based on the forward motion information and/or the backward motion information of the first peripheral block comprises:

determining the motion information of the first stage as forward motion information of the first surrounding block if the forward motion information of the first surrounding block is available and the forward motion information of the first surrounding block points to a first frame in a List 0;

if the forward motion information of the first surrounding block is available but the forward motion information of the first surrounding block does not point to the first frame in List0, scaling the forward motion information of the first surrounding block to point to the first frame in List0, and determining the motion vector of the first stage as a scaled motion vector, with the reference frame index being the index of the first frame in List 0;

If the forward motion information of the first surrounding block is not available, but the backward motion information is available, scaling the backward motion information of the first surrounding block to point to a first frame in List0, and determining that the motion vector of the first stage is a scaled motion vector, and the reference frame index is an index of the first frame in List 0;

the reference direction of the motion information in the first stage is the List0 direction.

6. The method of claim 3, wherein determining the motion information of the first stage based on the forward motion information and/or the backward motion information of the first peripheral block comprises:

If the backward motion information of the first peripheral block is not available, determining that the motion vector of the first stage is 0, and the reference frame index is the index of the first frame in List 1;

the reference direction of the motion information in the first stage is the List1 direction.

7. The method of claim 3, wherein determining the motion information of the first stage based on the forward motion information and/or the backward motion information of the first peripheral block comprises:

determining that the motion vector of the first stage is 0 and the reference frame index is an index of a first frame in List0 if the forward motion information of the first peripheral block is not available;

The reference direction of the motion information in the first stage is the List0 direction.

8. The method of claim 2, wherein determining the motion information of the first phase based on the first peripheral block comprises:

determining that the motion vector of the first stage is 0, the reference frame index is an index of a first frame in List0, and the reference direction of the motion information of the first stage is a List0 direction;

or the like, or, alternatively,

it is determined that the motion vector of the first stage is 0, the reference frame index is an index of a first frame in List1, and the reference direction of the motion information of the first stage is a List1 direction.

9. The method of any of claims 2-8, wherein determining the matching block for the current block based on the motion information of the first stage comprises:

and determining a matching block of the current block based on the position of the current block, the horizontal motion vector, the vertical motion vector and the precision of the motion vector in the first stage.

10. The method of any of claims 2-8, wherein determining the matching block for the current block based on the motion information of the first stage comprises:

and determining a matching block of the current block based on the position of the current block, the horizontal motion vector and the vertical motion vector of the first stage, the precision of the motion vector and the size of the sub-block.

11. The method according to any of claims 1-10, wherein a new matching block, obtained by offsetting the matching block, is determined by:

pruning the first sub-block and the second sub-block to be within the range of the current coding tree unit CTU, and comparing the motion information of the pruned first sub-block and second sub-block; and pruning the third sub-block and the fourth sub-block to be within the range of the current CTU, comparing the motion information of the pruned third sub-block and fourth sub-block, and if at least one of the two comparison results is different in motion information, horizontally shifting the matching block by one unit to the right to obtain a new matching block;

pruning the fifth sub-block and the sixth sub-block to be within the range of the current CTU, and comparing the motion information of the fifth sub-block and the sixth sub-block after pruning; and pruning the seventh sub-block and the eighth sub-block to be within the range of the current CTU, comparing the motion information of the seventh sub-block and the eighth sub-block after pruning, and if at least one of the two comparison results is different in motion information, horizontally shifting the matching block by one unit to the left to obtain a new matching block;

Pruning the first sub-block and the ninth sub-block to be within the range of the current CTU, and comparing the motion information of the pruned first sub-block and ninth sub-block; and pruning the fifth sub-block and the tenth sub-block to be within the range of the current CTU, comparing the motion information of the fifth sub-block and the tenth sub-block after pruning, and if at least one comparison result of the two comparison results is different in motion information, vertically and downwards offsetting the matching block by one unit to obtain a new matching block;

pruning the third sub-block and the eleventh sub-block to be within the range of the current CTU, and comparing the motion information of the pruned third sub-block and eleventh sub-block; and pruning the seventh sub-block and the twelfth sub-block to be within the range of the current CTU, comparing the motion information of the seventh sub-block and the twelfth sub-block after pruning, and if at least one of the two comparison results is different in motion information, vertically and upwardly offsetting the matching block by one unit to obtain a new matching block;

the first sub-block is the sub-block at the upper left corner of the matching block, the second sub-block is the adjacent sub-block at the top right corner of the matching block, the third sub-block is the sub-block at the lower left corner of the matching block, the fourth sub-block is the adjacent sub-block at the bottom right corner of the matching block, the fifth sub-block is the sub-block at the upper right corner of the matching block, the sixth sub-block is the adjacent sub-block at the top left corner of the matching block, the seventh sub-block is the sub-block at the lower right corner of the matching block, the eighth sub-block is the adjacent sub-block at the bottom left corner of the matching block, the ninth sub-block is the adjacent sub-block at the left corner of the matching block, the tenth sub-block is the adjacent sub-block at the right corner of the matching block, the eleventh sub-block is the adjacent sub-block at the left corner of the matching block, and the twelfth sub-block is the adjacent sub-block at the top right corner of the matching block; one unit is the side length of the subblock.

12. The method according to any of claims 1-10, wherein a new matching block, obtained by offsetting the matching block, is determined by:

and respectively carrying out horizontal direction and vertical direction deviation on the matching blocks based on one or more deviation amount pairs to obtain one or more new matching blocks.

13. The method of any of claims 1-10, wherein after determining the matching block for the current block, further comprising:

pruning the matching block to be within the range of the current CTU.

14. The method of claim 13, wherein a new matching block obtained by offsetting the matching block is determined by:

when the right boundary of the trimmed matching block is not located at the right boundary position of the current CTU, comparing the motion information of the thirteenth sub-block and the fourteenth sub-block, and comparing the motion information of the fifteenth sub-block and the sixteenth sub-block, if at least one of the two comparison results is that the motion information is different, horizontally shifting the trimmed matching block by one unit to the right to obtain a new matching block;

when the left boundary of the trimmed matching block is not located at the left boundary position of the current CTU, comparing the motion information of the seventeenth sub-block and the eighteenth sub-block, and comparing the motion information of the nineteenth sub-block and the twentieth sub-block, if at least one of the two comparison results is that the motion information is different, horizontally shifting the trimmed matching block by one unit to the left to obtain a new matching block;

When the lower boundary of the trimmed matching block is not located at the position of the lower boundary of the current CTU, comparing the motion information of the thirteenth sub-block with the motion information of the twenty-first sub-block, and comparing the motion information of the seventeenth sub-block with the motion information of the twenty-second sub-block, if at least one of the two comparison results is that the motion information is different, vertically and downwards shifting the trimmed matching block by one unit to obtain a new matching block;

when the upper boundary of the trimmed matching block is not located at the upper boundary position of the current CTU, comparing the motion information of the fifteenth sub-block with the motion information of the twenty-third sub-block, and comparing the motion information of the nineteenth sub-block with the motion information of the twenty-fourth sub-block, if at least one of the two comparison results is that the motion information is different, vertically and upwardly shifting the trimmed matching block by one unit to obtain a new matching block;

wherein the thirteenth sub-block is the sub-block at the top left corner of the trimmed matching block, the fourteenth sub-block is the adjacent sub-block at the top right corner of the trimmed matching block, the fifteenth sub-block is the sub-block at the bottom left corner of the trimmed matching block, the sixteenth sub-block is the adjacent sub-block at the bottom right corner of the trimmed matching block, the seventeenth sub-block is the sub-block at the top right corner of the trimmed matching block, the eighteenth sub-block is the adjacent sub-block at the top left corner of the trimmed matching block, the nineteenth sub-block is the sub-block at the bottom right corner of the trimmed matching block, the twentieth sub-block is the adjacent sub-block at the bottom left corner of the trimmed matching block, the twenty-first sub-block is the adjacent sub-block at the left corner below the trimmed matching block, the twenty-second sub-block is the adjacent sub-block at the right corner below the trimmed matching block, the twenty-third sub-block is an adjacent sub-block on the leftmost side right above the trimmed matching block, and the twenty-fourth sub-block is an adjacent sub-block on the rightmost side right above the trimmed matching block; one unit is the side length of the subblock.

15. The method according to any of claims 11-14, wherein said determining a candidate enhanced temporal motion vector prediction mode based on said matching block and a new matching block obtained by shifting said matching block comprises:

when at least one new matching block exists, determining a prediction mode corresponding to the matching block before shifting and a prediction mode corresponding to the new matching block obtained by shifting as a candidate enhanced temporal motion vector prediction mode;

or the like, or, alternatively,

when at least one new matching block exists, determining a prediction mode corresponding to the new matching block obtained by shifting as a candidate enhanced temporal motion vector prediction mode;

or the like, or, alternatively,

and when no new matching block exists, determining the prediction mode corresponding to the matching block before shifting as a candidate enhanced temporal motion vector prediction mode.

16. The method according to any of claims 1-12 or 15, wherein said determining motion information for each sub-block in the current block based on the enhanced temporal motion vector prediction mode comprises:

for any sub-block in the target matching block, pruning the sub-block to be within the range of the current CTU; the target matching block is a matching block corresponding to the enhanced temporal motion vector prediction mode;

If the forward motion information and the backward motion information of the pruned subblock are both available, respectively scaling the forward motion information and the backward motion information of the pruned subblock to point to a first frame of List0 and a first frame of List1, and respectively endowing the scaled forward motion information and backward motion information to the subblock at the corresponding position of the current block;

if the forward motion information of the pruned sub-block is available but the backward motion information is not available, scaling the forward motion information of the pruned sub-block to a first frame pointing to List0, and giving the scaled forward motion information to the sub-block at the corresponding position of the current block;

if the backward motion information of the pruned sub-block is available but the forward motion information is not available, the backward motion information of the pruned sub-block is scaled to the first frame pointing to the List1, and the scaled backward motion information is assigned to the sub-block at the corresponding position of the current block.

17. The method of claim 16, wherein determining motion information for each sub-block in the current block based on the enhanced temporal motion vector prediction mode further comprises:

if the forward motion information and the backward motion information of the pruned subblock are unavailable, pruning the center position of the target matching block to be within the range of the current CTU, respectively stretching the forward motion information and the backward motion information of the center position of the pruned target matching block to a first frame pointing to List0 and a first frame pointing to List1 when both the forward motion information and the backward motion information of the center position of the pruned target matching block are available, and respectively endowing the stretched forward motion information and the stretched backward motion information to the subblock at the corresponding position of the current block; when the forward motion information of the center position of the pruned target matching block is available but the backward motion information is not available, scaling the forward motion information of the center position of the pruned target matching block to a first frame pointing to a List0, and assigning the scaled forward motion information to a sub-block at a corresponding position of the current block; when the backward motion information of the center position of the pruned target matching block is available but the forward motion information is not available, scaling the backward motion information of the center position of the pruned target matching block to a first frame pointing to a List1, and assigning the scaled backward motion information to a sub-block at a corresponding position of the current block; when the forward motion information and the backward motion information of the center position of the trimmed target matching block are unavailable, giving zero motion information to the sub-block at the position corresponding to the current block;

Or if the forward motion information and the backward motion information of the pruned sub-blocks are unavailable, giving zero motion information to the sub-block at the corresponding position of the current block;

or, if the forward motion information and the backward motion information of the pruned subblock are unavailable, when the forward motion information and the backward motion information of the second surrounding block of the current block are both available, respectively scaling the forward motion information and the backward motion information of the second surrounding block to a first frame pointing to List0 and a first frame pointing to List1, and respectively assigning the scaled forward motion information and the scaled backward motion information to the subblock at the corresponding position of the current block; when the forward motion information of the second bounding box is available but the backward motion information is not available, scaling the forward motion information of the second bounding box to a first frame pointing to List0, and giving the scaled forward motion information to the sub-block at the corresponding position of the current block; when the backward motion information of the second surrounding block is available but the forward motion information is not available, scaling the backward motion information of the second surrounding block to a first frame pointing to List1, and assigning the scaled backward motion information to the sub-block at the corresponding position of the current block; and when the forward motion information and the backward motion information of the second surrounding block are unavailable, giving zero motion information to the sub-block at the corresponding position of the current block.

18. The method according to any of claims 1-15, wherein said determining motion information for each sub-block in said current block based on said enhanced temporal motion vector prediction mode comprises:

for any sub-block in the target matching block, if the forward motion information and the backward motion information of the sub-block are both available, respectively scaling the forward motion information and the backward motion information of the sub-block to point to the first frame of List0 and the first frame of List1, and respectively giving the scaled forward motion information and backward motion information to the sub-block at the corresponding position of the current block;

if the forward motion information of the sub-block is available, but the backward motion information is not available, the forward motion information of the sub-block is extended to the first frame pointing to the List0, and the extended forward motion information is given to the sub-block at the corresponding position of the current block;

if the backward motion information of the sub-block is available but the forward motion information is not available, the backward motion information of the sub-block is scaled to the first frame pointing to the List1, and the scaled backward motion information is assigned to the sub-block at the corresponding position of the current block.

19. The method of claim 18, wherein determining motion information for each sub-block in the current block based on the enhanced temporal motion vector prediction mode further comprises:

If the forward motion information and the backward motion information of the subblock are unavailable, respectively scaling the forward motion information and the backward motion information of the central position of the target matching block to a first frame pointing to List0 and a first frame pointing to List1 when the forward motion information and the backward motion information of the central position of the target matching block are both available, and respectively assigning the scaled forward motion information and the scaled backward motion information to the subblock at the corresponding position of the current block; when the forward motion information of the center position of the target matching block is available, but the backward motion information is not available, scaling the forward motion information of the center position of the target matching block to a first frame pointing to the List0, and assigning the scaled forward motion information to the sub-block at the corresponding position of the current block; when the backward motion information of the center position of the target matching block is available but the forward motion information is not available, scaling the backward motion information of the center position of the target matching block to a first frame pointing to the List1, and assigning the scaled backward motion information to the sub-block at the corresponding position of the current block; when the forward motion information and the backward motion information of the central position of the target matching block are unavailable, giving zero motion information to the subblock at the corresponding position of the current block;

Or if the forward motion information and the backward motion information of the sub-block are unavailable, giving zero motion information to the sub-block at the position corresponding to the current block;

or, if the forward motion information and the backward motion information of the subblock are unavailable, when the forward motion information and the backward motion information of the second surrounding block of the current block are available, respectively scaling the forward motion information and the backward motion information of the second surrounding block to a first frame pointing to List0 and a first frame pointing to List1, and respectively assigning the scaled forward motion information and backward motion information to the subblock at the corresponding position of the current block; when the forward motion information of the second bounding box is available but the backward motion information is not available, scaling the forward motion information of the second bounding box to a first frame pointing to List0, and giving the scaled forward motion information to the sub-block at the corresponding position of the current block; when the backward motion information of the second surrounding block is available but the forward motion information is not available, scaling the backward motion information of the second surrounding block to a first frame pointing to List1, and assigning the scaled backward motion information to the sub-block at the corresponding position of the current block; and when the forward motion information and the backward motion information of the second surrounding block are unavailable, giving zero motion information to the sub-block at the corresponding position of the current block.

20. The method according to any of claims 1-19, wherein said parsing index information of enhanced temporal motion vector prediction mode from the bitstream of the current block comprises:

and when the current block is determined to enable the enhanced time domain motion vector prediction technology, analyzing index information of an enhanced time domain motion vector prediction mode from a code stream of the current block.

21. The method of claim 20, wherein whether the current block activates an enhanced temporal motion vector prediction technique is indicated using a sequence parameter set level syntax or a slice level syntax.

22. The method of claim 21, wherein whether the current block enables an enhanced temporal motion vector prediction technique is determined by, when the current block enables an enhanced temporal motion vector prediction technique using sequence parameter set level syntax:

when the image sequence to which the current block belongs enables the enhanced temporal motion vector prediction technology, determining that the current block enables the enhanced temporal motion vector prediction technology;

when the image sequence to which the current block belongs does not enable the enhanced temporal motion vector prediction technology, determining that the current block does not enable the enhanced temporal motion vector prediction technology.

23. The method of claim 21, wherein when using a slice-level syntax to indicate whether the current block enables an enhanced temporal motion vector prediction technique, whether the current block enables the enhanced temporal motion vector prediction technique is determined by:

determining that the current block enables an enhanced temporal motion vector prediction technique when the slice to which the current block belongs enables the enhanced temporal motion vector prediction technique;

when the enhanced temporal motion vector prediction technique is not enabled for the slice to which the current block belongs, determining that the enhanced temporal motion vector prediction technique is not enabled for the current block.

24. The method of claim 20, wherein whether the current block enables an enhanced temporal motion vector prediction technique is determined by:

when the size of the current block is smaller than or equal to the size of a preset maximum coding block and larger than or equal to the size of a preset minimum coding block, determining that the current block enables an enhanced time domain motion vector prediction technology;

and when the size of the current block is larger than the size of a preset maximum coding block or smaller than the size of a preset minimum coding block, determining that the enhanced temporal motion vector prediction technology is not started for the current block.

25. The method of claim 24, wherein the size of the preset largest coding block is expressed using a sequence parameter set level syntax or using a slice level syntax;

and/or the first and/or second light sources,

the size of the preset minimum coding block is represented by using a sequence parameter set level syntax or by using a slice level syntax.

26. A method of encoding, comprising:

determining a matching block of a current block based on a first peripheral block of the current block;

determining a candidate enhanced temporal motion vector prediction mode based on the matching block and a new matching block obtained by shifting the matching block, and constructing a first temporal candidate mode list based on the candidate enhanced temporal motion vector prediction mode;

traversing each candidate enhanced temporal motion vector prediction mode in the first temporal candidate mode list, determining motion information of each subblock in the current block based on any candidate enhanced temporal motion vector prediction mode and the candidate enhanced temporal motion vector prediction mode, and performing motion compensation on each subblock in the current block based on the motion information of each subblock in the current block;

determining the candidate enhanced temporal motion vector prediction mode with the minimum rate distortion cost as the enhanced temporal motion vector prediction mode of the current block based on the rate distortion cost corresponding to each candidate enhanced temporal motion vector prediction mode;

And carrying index information of the enhanced temporal motion vector prediction mode of the current block in a code stream of the current block, wherein the index information is used for identifying the position of the enhanced temporal motion vector prediction mode in the first temporal candidate mode list.

27. A decode-side device comprising a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor for executing the machine-executable instructions to implement the method of any one of claims 1-25.

28. An encoding end device comprising a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor, the processor being configured to execute the machine-executable instructions to implement the method of claim 26.

Technical Field

The present application relates to video encoding and decoding technologies, and in particular, to a decoding method, an encoding method, an apparatus, and a device.

Background

Complete video coding generally includes operations of prediction, transformation, quantization, entropy coding, filtering, and so on. The prediction can be divided into intra-frame prediction and inter-frame prediction, wherein the intra-frame prediction is to predict a current uncoded block by using surrounding coded blocks as references, and effectively remove redundancy on a spatial domain. Inter-frame prediction is to use neighboring coded pictures to predict the current picture, effectively removing redundancy in the temporal domain.

An optional Temporal Motion Vector Prediction (ATMVP) technique adopted in the general Video Coding (VVC) standard is to provide different Motion information for each subblock in a current Coding unit by using Motion information of a Temporal subblock, and the method is to find a Coding block corresponding to an adjacent block of a current Coding block in a Co-located (common position) frame according to Motion information of the adjacent block of the current Coding block, and then provide Motion information of each subblock inside the Coding block at a position corresponding to the current Coding block in the Co-located frame to each subblock in the current Coding block.

However, in the conventional ATMVP technology, finding a coding block in a co-located frame corresponding to a current coding block needs to depend on motion information of surrounding blocks of the current coding block, and if the motion information of the surrounding blocks is inaccurate, the motion information of the found coding block in the corresponding position is unreliable, thereby affecting the coding performance.

Disclosure of Invention

In view of the above, the present application provides a decoding method, an encoding method, an apparatus and a device.

Specifically, the method is realized through the following technical scheme:

according to a first aspect of embodiments of the present application, there is provided a decoding method, including:

determining a matching block of the current block based on a first surrounding block of the current block;

determining an enhanced temporal motion vector prediction mode from the second temporal candidate mode list based on the index information;

In some embodiments, said determining a matching block for the current block based on a first surrounding block of the current block comprises:

determining motion information of a first stage based on the first peripheral block;

determining a matching block for the current block based on the motion information of the first stage.

In some embodiments, the determining motion information of the first stage based on the first peripheral block includes:

determining motion information of a first stage based on the forward motion information and/or the backward motion information of the first peripheral block.

In some embodiments, the determining motion information of the first stage based on the forward motion information and/or backward motion information of the first peripheral block includes: