Motion vector prediction method and related apparatus

Patent metadata

Field	Value
Publication number	US-2021127116-A1
Application number	US-202117140041-A
Country	US
Kind code	A1
Filing date	Jan 1, 2021
Priority date	Jul 2, 2018
Publication date	Apr 29, 2021
Grant date	—

Abstract

A method includes: parsing a bitstream to obtain an index value of a candidate motion vector list; constructing the candidate motion vector list, where the candidate motion vector list includes candidate motion vectors of K control points of a current block, the candidate motion vectors of the K control points are obtained based on a 2N-parameter affine transform model used for a neighboring block of the current block, N, K are integers greater than or equal to 2 and less than or equal to 4, and N is not equal to K; determining, in the candidate motion vector list, target candidate motion vectors of the K control points based on the index value; and obtaining a predicted motion vector of each subblock of the current block based on the target candidate motion vectors of the K control points. It helps improve coding efficiency in an audio and video coding process.

First claim

What is claimed is:

1 . A motion vector prediction method, wherein the method comprises:
parsing a bitstream to obtain an index value of a candidate motion vector list; when an affine model of a current block is a 2×K-parameter affine transform model, and an affine transform model of a neighboring block of the current block is a 2N-parameter affine transform model, obtaining, according to the 2×N-parameter affine transform model used for the neighboring block, candidate motion vectors of K control points of the current block, wherein the 2×N-parameter affine transform model is obtained based on motion vectors of N control points of the neighboring block, N is an integer greater than or equal to 2 and less than or equal to 4, K is an integer greater than or equal to 2 and less than or equal to 4, N is not equal to K, the neighboring block is a decoded image block spatially adjacent to the current block, and the current block comprises a plurality of subblocks; constructing the candidate motion vector list, wherein the candidate motion vector list comprises candidate motion vectors of K control points of the current block; determining, in the candidate motion vector list, target candidate motion vectors of the K control points based on the index value; obtaining a predicted motion vector of each subblock of the plurality of subblocks in the current block based on the target candidate motion vectors of the K control points; and generating a prediction block of the current block based on the predicted motion vectors of the plurality of subblocks in the current block.

2 . The method according to claim 1 , wherein N is equal to 2 and K is equal to 3, and correspondingly, candidate motion vectors of three control points of the current block are obtained based on a 4-parameter affine transform model used for the neighboring block of the current block.

3 . The method according to claim 2 , wherein the candidate motion vectors of the three control points of the current block comprise a motion vector corresponding to a top-left sample location of the current block, a motion vector corresponding to a top-right sample location of the current block, and a motion vector corresponding to a bottom-left sample location of the current block; and
that the candidate motion vectors of the three control points of the current block are obtained based on the 4-parameter affine transform model used for the neighboring block of the current block comprises that the candidate motion vectors of the three control points of the current block are calculated according to the following formulas:

{

vx
0

=

vx
4

+

(

vx
5

-

vx
4

)

x
5

-

x
4

×

(

x
0

-

x
4

)

-

(

vy
5

-

vy
4

)

x
5

-

x
4

×

(

y
0

-

y
4

)

vy
0

=

vx
4

+

(

vy
5

-

vy
4

)

x
5

-

x
4

×

(

Assignees

• Huawei Tech Co Ltd

Inventors

• Chen Huanbang
• Yang Haitao
• Chen Jianle
• Fu Jiali

Classifications

• H04N19/52

  by predictive encoding · CPC title

• H04N19/54 Primary

  using feature points or meshes · CPC title

• H04N19/119

  Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title

• H04N19/139 Primary

  Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability · CPC title