Unified merge mode and adaptive motion vector prediction mode candidates selection

• US 9,282,338 B2
• Filed: 06/04/2012
• Issued: 03/08/2016
• Est. Priority Date: 06/20/2011
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Zheng, Yunfei, Wang, Xianglin, Karczewicz, Marta
• Primary Examiner(s): Czekaj, Dave
• Assistant Examiner(s): KALAPODAS, DRAMOS
• Application Number:
  US13/487,996
• Publication Number:
  US 20120320969A1
• Time in Patent Office: 1,373 Days
• Field of Search: 375/240.02
• US Class Current: 1/1
• CPC Class Codes: H04N 19/40 using video transcoding, i.e. partial or full decoding of a coded input stream followed by re-encoding of the decoded output stream
• CPC Class Codes: H04N 19/52 by predictive encoding

Abstract

A unified candidate block set for both adaptive motion vector prediction (AMVP) mode and merge mode for use in inter-prediction is proposed. In general, the same candidate block set is used regardless of which motion vector prediction mode (e.g., merge mode or AMVP mode) is used. In other examples of this disclosure, one candidate block in a set of candidate blocks is designated as an additional candidate block. The additional candidate block is used if one of the other candidate blocks is unavailable. Also, the disclosure proposes a checking pattern where the left candidate block is checked before the below left candidate block. Also, the above candidate block is checked before the right above candidate block.

Claims

1. A method of encoding a motion vector in a video coding process, the method comprising:

determining one of a plurality of modes for a motion vector prediction process for a current block of video data, the determined mode being a merge mode;

determining a set of candidate blocks, the determined set comprising a below-left candidate block, a left candidate block, an above candidate block, a right-above candidate block, a left-above candidate block, and a temporal candidate block;

performing the motion vector prediction process for the current block using the determined mode and the set of candidate blocks, and

wherein one candidate block in the set of candidate blocks is designated as an additional candidate block, and wherein the additional candidate block is used in the merge mode if another of the candidate blocks of the set of candidate blocks is unavailable or invalid.

2. The method of claim 1, further comprising:

determining a motion vector relative to a reference frame for the current block of video data, wherein a location of a reference frame is based on a prediction direction.

3. The method of claim 2, the merge mode having a maximum number N of candidate blocks for use in performing the motion vector prediction process,

wherein the motion vector prediction process is performed according to a checking pattern, the checking pattern defining an order for checking each of candidate blocks in the set of candidate blocks, and

wherein the set of candidate blocks is defined as the first N available candidate blocks in the set of candidate blocks along the checking pattern.

4. The method of claim 3, wherein the checking pattern is based on one or more of a block size, a partition size, and a partition index.

5. The method of claim 3, further comprising:

updating for the checking pattern for subsequently coded blocks based on one or more of a merge index, the determined mode, a partition size, a partition index, a reference frame index, a motion vector difference, and a motion vector prediction.

6. The method of claim 1, wherein the set of candidate blocks include a below left candidate block, a left candidate block, an above candidate block, a right above candidate block, a left above candidate block, and a temporal candidate block, and wherein the additional candidate block is the left above candidate block.

7. The method of claim of claim 1, wherein the additional candidate block is a candidate block that is in a causal relationship to the current block.

8. An apparatus configured to encode a motion vector in a video coding process comprising:

a memory configured to store motion vector information for a set of candidate blocks; and

a video encoder configured to;

determine one of a plurality of modes for a motion vector prediction process for a current block of video data, the determined mode being a merge mode;

determine a set of candidate blocks, the determined set comprising a below-left candidate block, a left candidate block, an above candidate block, a fight-above candidate block, a left-above candidate block, and a temporal candidate block;

perform the motion vector prediction process for the current block using the determined mode and the set of candidate blocks; and

wherein one candidate block in the set of candidate blocks is designated as an additional candidate block, and wherein the additional candidate block is used in the merge mode if another of the candidate blocks of the set of candidate blocks is unavailable or invalid.

9. The apparatus of claim 8, wherein the video encoder is further configured to:

determine a motion vector relative to a reference frame for the current block of video data, wherein a location of a reference frame is based on a prediction direction.

10. The apparatus of claim 9, the merge mode having a maximum number N of candidate blocks for use in performing the motion vector prediction process,

wherein the motion vector prediction process is performed according to a checking pattern, the checking pattern defining an order for checking each of candidate blocks in the set of candidate blocks;

and

wherein the set of candidate blocks is defined as the first N available candidate blocks in the set of candidate blocks along the checking pattern.

11. The apparatus of claim 10, wherein the checking pattern is based on one or more of a block size, a partition size, and a partition index.

12. The apparatus of claim 10, wherein the video encoder is further configured to:

update the checking pattern for subsequently coded blocks based on one or more of a merge index, the determined mode, a partition size, a partition index, a reference frame index, a motion vector difference, and a motion vector prediction.

13. The apparatus of claim 8, wherein the set of candidate blocks include a below left candidate block, a left candidate block, an above candidate block, a right above candidate block, a left above candidate block, and a temporal candidate block, and wherein the additional candidate block is the left above candidate block.

14. The apparatus of claim of claim 8, wherein the additional candidate block is a candidate block that is in a causal relationship to the current block.

15. A non-transitory computer-readable storage medium having stored thereon instructions that, when executed, cause a processor of a device for encoding video data to:

determine one of a plurality of modes for a motion vector prediction process for a current block of video data, the determined mode being a merge mode an;

determine a set of candidate blocks, the determined set comprising a below-left candidate block, a left candidate block, an above candidate block, a right-above candidate block, a left- above candidate block, and a temporal candidate block;

perform the motion vector prediction process for the current block using the determined mode and the set of candidate blocks;

wherein one candidate block in the set of candidate blocks is designated as an additional candidate block, and wherein the additional candidate block is used in the merge mode if another of the candidate blocks of the set of candidate blocks is unavailable or invalid.

16. A method of decoding a motion vector in a video coding process comprising:

determining one of a plurality of modes for a motion vector prediction process for a current block of video data, the determined mode being a merge mode;

determining a candidate block from a set of candidate blocks comprising a below-left candidate block, a left candidate block, an above candidate block, a fight-above candidate block, a left-above candidate block, and a temporal candidate block; and

wherein one candidate block in the set of candidate blocks is designated as an additional candidate block, wherein the additional candidate block is used in the merge mode if another of the candidate blocks of the set of candidate blocks is unavailable or invalid, and wherein information associated with the candidate block is used to decode a motion vector for the current block.

17. The method of claim 16, wherein the mode for the motion vector prediction process is determined from a received syntax element, and wherein the candidate block is determined from a received syntax element.

18. The method of claim 16, wherein the method further comprises:

retrieving a motion vector, a reference frame, and a prediction direction associated with the candidate block having the received index; and

performing an inter-prediction process on the current block using the retrieved motion vector, reference frame, and prediction direction.

19. The method of claim 16, the merge mode having a maximum number N of candidate blocks for use in performing the motion vector prediction process,

wherein the motion vector prediction process is performed according to a checking pattern, the checking pattern defining an order for checking each of candidate blocks in the set of candidate blocks;

and

wherein the set of candidate blocks is defined as the first N available candidate blocks in the set of candidate blocks along the checking pattern.

20. The method of claim 19, wherein the checking pattern is based on one or more of a block size, a partition size, and a partition index.

21. The method of claim 19, further comprising:

updating the checking order pattern for subsequently coded blocks based on one or more of a merge index, the determined mode, a partition size, a partition index, a reference frame index, a motion vector difference, and a motion vector prediction.

22. The method of claim 16, wherein the set of candidate blocks include a below left candidate block, a left candidate block, an above candidate block, a right above candidate block, a left above candidate block, and a temporal candidate block, and wherein the additional candidate block is the left above candidate block.

23. The method of claim of claim 16, wherein the additional candidate block is a candidate block that is in a causal relationship to the current block.

24. An apparatus configured to decode a motion vector in a video coding process comprising, the apparatus comprising:

a memory configured to store motion vector information for a set of candidate blocks; and

a video decoder configured to;

determine one of a plurality of modes for a motion vector prediction process for a current block of video data, the determined mode being a merge mode;

determine a candidate block from the set of candidate blocks comprising a below- left candidate block, a left candidate block, an above candidate block, a fight-above candidate block, a left-above candidate block, and a temporal candidate block;

wherein if the one candidate block in the set of candidate blocks is designated as an additional candidate block, wherein the additional candidate block is used in the merge mode if another of the candidate blocks of the set of candidate blocks is unavailable or invalid, and wherein information associated with the candidate block is used to decode a motion vector for the current block.

25. The apparatus of claim 24, wherein the mode for the motion vector prediction process is determined from a received syntax element, and wherein the candidate block is determined from a received syntax element.

26. The apparatus of claim 24, wherein the video coder is further configured to:

retrieve a motion vector, a reference frame, and a prediction direction associated with the candidate block having the received index; and

perform an inter-prediction process on the current block using the retrieved motion vector, reference frame, and prediction direction.

27. The apparatus of claim 24, the merge mode having a maximum number N of candidate blocks for use in performing the motion vector prediction process,

wherein the motion vector prediction process is performed according to a checking pattern, the checking pattern defining an order for checking each of candidate blocks in the set of candidate blocks;

and

wherein the set of candidate blocks is defined as the first N available candidate blocks in the set of candidate blocks along the checking pattern.

28. The apparatus of claim 27, wherein the checking pattern is based on one or more of a block size, a partition size, and a partition index.

29. The apparatus of claim 27, wherein the video decoder is further configured to:

update the checking order pattern for subsequently coded blocks based on one or more of a merge index, the determined mode, a partition size, a partition index, a reference frame index, a motion vector difference, and a motion vector prediction.

30. The apparatus of claim 24, wherein the set of candidate blocks include a below left candidate block, a left candidate block, an above candidate block, a right above candidate block, a left above candidate block, and a temporal candidate block, and wherein the additional candidate block is the left above candidate block.

31. The apparatus of claim of claim 24, wherein the additional candidate block is a candidate block that is in a causal relationship to the current block.

32. A non-transitory computer-readable storage medium having stored thereon instructions that, when executed, cause a processor of a device for decoding video data to:

determine one of a plurality of modes for a motion vector prediction process for a current block of video data, the determined mode being a merge mode;

determine a candidate block from a set of candidate blocks comprising a below-left candidate block, a left candidate block, an above candidate block, a right-above candidate block, a left-above candidate block, and a temporal candidate block;

wherein one candidate block in the set of candidate blocks is designated as an additional candidate block, wherein the additional candidate block is used in the merge mode if another of the candidate blocks of the set of candidate blocks is unavailable or invalid, and wherein information associated with the candidate block is used to decode a motion vector for the current block.