Method of Motion Vector Derivation for Video Coding

• US 20170214932A1
• Filed: 07/15/2015
• Published: 07/27/2017
• Est. Priority Date: 07/18/2014
• Status: Active Grant

• Current Assignee: HFI Innovation, Inc. (MediaTek, Inc.)
• Original Assignee: MediaTek Singapore Pte. Ltd. (MediaTek, Inc.)
• Inventors: HUANG, Han
• Granted Patent: US 10,582,210 B2
• Time in Patent Office: Days
• CPC Class Codes: H04N 19/46 Embedding additional information in the video signal during the compression process H04N19/517, H04N19/68, H04N19/70 take precedence
• CPC Class Codes: H04N 19/513 Processing of motion vectors
• CPC Class Codes: H04N 19/52 by predictive encoding
• CPC Class Codes: H04N 19/527 Global motion vector estimation
• CPC Class Codes: H04N 19/537 Motion estimation other than block-based
• CPC Class Codes: H04N 19/70 characterised by syntax aspects related to video coding, e.g. related to compression standards
• CPC Class Codes: H04N 19/96 Tree coding, e.g. quad-tree coding

Abstract

A method and apparatus for deriving a sub-block motion vector for the current sub-block based on a motion-model function depending on the current sub-block location are disclosed. The derived sub-block motion vector is then used for encoding or decoding the sub-block. The motion-model function may correspond to an affine motion-model function or a bilinear motion-model function. In one embodiment, a new Merge mode can be used to apply prediction of a current block by applying prediction on the sub-block basis using the sub-block motion vector derived from the motion-model function. In another embodiment, an additional inter prediction mode can be used to apply prediction of a current block by applying prediction on the sub-block basis using the sub-block motion vector derived from the motion-model function.

Claims

1. A method of motion information derivation for blocks of a picture, the method comprising:

receiving input data associated with a current block;

determining a current sub-block located at a current sub-block location within the current block;

calculating a derived sub-block motion vector for the current sub-block based on a motion-model function depending on the current sub-block location; and

encoding or decoding the current sub-block using the derived sub-block motion vector for the current sub-block.

2. The method claim 1, wherein the motion-model function corresponds to an affine motion-model function or a bilinear motion-model function.

3. The method claim 2, wherein the motion-model function corresponds to the affine motion-model function represented by F(x,y)=a×

x+b×

y+e, and wherein (x,y) is the current sub-block location and a, b and e are affine model parameters associated with the affine motion-model function.

4. The method claim 3, wherein the current block has a block size of M×

N and three affine parameter motion vectors, MV_0, MV_1 and MV_2 are determined for locations (0,

     0), (M,

     0) and (0, N) respectively, and the affine model parameters are determined according to e=MV_0, a=(MV_1−

MV_0)/M and b=(MV_2−

MV_0)/N, and wherein M and N are positive integers.

5. The method claim 4, wherein the three affine parameter motion vectors are derived based on neighboring coded blocks of the current block.

6. The method claim 5, wherein if an affine Merge mode using the three affine parameter motion vectors is enabled, the three affine parameter motion vectors are inserted into a Merge candidate list for encoding or decoding of the current block, wherein one sub-block motion vector is derived for each sub-block of the current block based on the affine motion-model function and the three affine parameter motion vectors, and wherein each sub-block is predicted using said one sub-block motion vector if the affine Merge mode is used for the current block.

7. The method claim 6, wherein one or more syntax elements are used to indicate whether the affine Merge mode is enabled, positions of the three affine parameter motion vectors in the Merge candidate list, or both.

9. The method claim 7, wherein the affine Merge mode is enabled only for the blocks of the picture belonging to a selected PU partition type set, wherein the selected PU partition type set consists of one or more PU partition types selected from {2N×

2N, 2N×

N, N×

2N, 2N×

nU, 2N×

nD, nL×

2N, nR×

2N}.

10. The method claim 7, wherein said one or more syntax elements are derived implicitly at a decoder side.

11. The method claim 10, wherein said one or more syntax elements are implicitly derived according to coding mode selections, motion parameters of neighboring blocks of the current block, or motion parameters of reference blocks of the current bock.

12. The method claim 4, wherein one or more of the three affine parameter motion vectors, MV_0, MV_1 and MV_2 are determined by refining respective one or more MV predictors to form one or more refined motion vectors using one or more transmitted motion vector differences (MVDs).

13. The method claim 12, wherein if an additional Inter prediction mode is enabled, one sub-block motion vector is derived for each sub-block of the current block based on the motion-model function and the three affine parameter motion vectors after said one or more of the three affine parameter motion vectors being refined, and each sub-block is predicted using said one sub-block motion vector if the additional Inter prediction mode is used for the current block.

14. The method claim 13, wherein a syntax element is used to indicate whether the additional Inter prediction mode is enabled.

15. The method claim 14, wherein the syntax element is signaled after a second syntax element for a Merge flag for the current block in a bitstream, and wherein the Merge flag indicates whether the current block is coded using a Merge mode.

16. The method claim 14, wherein the syntax element is derived implicitly at a decoder side.

17. The method claim 16, wherein the syntax element is implicitly derived according to coding mode selections, motion parameters of neighboring blocks of the current block, or motion parameters of reference blocks of the current bock.

19. The method claim 4, wherein the derived sub-block motion vector is quantized to ¼

, ⅛

, 1/12, 1/16 or 1/32 pixel accuracy.

20. The method claim 1, wherein the current block corresponds to a prediction unit (PU) and sub-block size corresponds to 4×

4, 8×

8 or 16×

16.

8. (canceled)

18. (canceled)

21. An apparatus for motion information derivation for blocks of a picture, the apparatus comprising one or more electronic circuits configure to:

receive input data associated with a current block;

determine a current sub-block located at a current sub-block location within the current block;

calculate a derived sub-block motion vector for the current sub-block based on a motion-model function depending on the current sub-block location; and

encode or decode the current sub-block using the derived sub-block motion vector for the current sub-block.