Image encoding/decoding method and apparatus for throughput enhancement using motion vector refinement

Patent metadata

Field	Value
Publication number	US-2025392746-A1
Application number	US-202519311316-A
Country	US
Kind code	A1
Filing date	Aug 27, 2025
Priority date	Jun 29, 2018
Publication date	Dec 25, 2025
Grant date	—

Abstract

There is provided an image encoding/decoding method and apparatus. The image decoding method comprises decoding information on whether or not to perform motion information refinement from a bitstream, obtaining second motion information for first motion information on a current block on the basis of the information and reconstructing at least one of the current block and a neighboring block of the current block by using at least one of the first motion information and the second motion information.

First claim

1 . A method of decoding an image, the method comprising:
constructing a motion vector candidate list including a spatial motion vector and a temporal motion vector, from a neighboring block in a current picture including a current block and a col-block of the current block in a different picture from the current picture; deriving a first motion vector of the current block using the motion vector candidate list; decoding information on whether or not to perform motion vector refinement from a bitstream; obtaining a second motion vector of the current block by refining a first motion vector of the current block based on the information on whether or not to perform motion vector refinement; generating a prediction block of the current block in an inter-prediction mode based on the second motion vector; generating a residual block of the current block from the bitstream; and reconstructing the current block based on the prediction block, wherein constructing the motion vector candidate list comprises:
entering a motion vector of the neighboring block as the spatial motion vector;
obtaining a refined motion vector of the col-block;
scaling the refined motion vector of the col-block; and
entering the scaled refined motion vector of the col-block as the temporal motion vector, and
wherein the motion vector of the neighboring block is excluded from the motion vector candidate list when the neighboring block is included in a region including the current block, the region being derived based on size information from the bitstream.

2 . The method of claim 1 , wherein the information on whether or not to perform motion vector refinement is signaled through at least one of a parameter set or video parameter set, a sequence parameter set, a subpicture parameter set, a picture parameter set, an adaptation parameter set, a slice header, a tile header, and a brick header.

3 . The method of claim 1 , wherein the second motion vector is obtained by partitioning the current block into at least one subblock, and performing motion vector refinement on the at least one subblock when at least one of a width and a height of the current block is greater than a predetermined block size.

4 . The method of claim 3 , wherein the predetermined block size is 16.

5 . The method of claim 3 , wherein an initial motion vector of the current block is used as an initial motion vector of the subblock.

6 . The method of claim 1 , wherein the second motion vector is obtained by using a pixel value indicated by a candidate motion vector in a region within an L0 reference picture for the current block, and using a pixel value indicated by a vector opposite to the candidate motion vector in a region within an L1 reference picture for the current block.

7 . The method of claim 1 , wherein the first motion vector of the current block is used for determining a boundary strength of deblocking filtering.

8 . A method of encoding an image, the method comprising:
constructing a motion vector candidate list including a spatial motion vector and a temporal motion vector, from a neighboring block in a current picture including a current block and a col-block of the current block in a different picture from the current picture; determining a first motion vector of the current block using the motion vector candidate list; determining whether or not to perform motion vector refinement; obtaining a second motion vector of the current block by refining a first motion vector of the current block based on the determination of whether or not to perform motion vector refinement; generating a prediction block of the current block in an inter-prediction mode based on the second motion vector; generating a residual block of the current block based on the prediction block; generating transform coefficients of the current block based on the residual block of the current block; and encoding information on whether or not to perform motion vector refinement based on the determination of whether or not to perform motion vector refinement, wherein constructing the motion vector candidate list comprises:
entering a motion vector of the neighboring block as the spatial motion vector;
obtaining a refined motion vector of the col-block;
scaling the refined motion vector of the col-block; and
entering the scaled refined motion vector of the col-block as the temporal motion vector, and
wherein the motion vector of the neighboring block is excluded from the motion vector candidate list when the neighboring block is included in a region including the current block, the region being derived based on size information from a bitstream.

9 . The method of claim 8 , wherein the information on whether or not to perform motion vector refinement is loaded in at least one of a parameter set or video parameter set, a sequence parameter set, a subpicture parameter set, a picture parameter set, an adaptation parameter set, a slice header, a tile header, and a brick header.

10 . The method of claim 8 , wherein the second motion vector is obtained by partitioning the current block into at least one subblock, and performing motion vector refinement on the at least one subblock when at least one of a width and a height of the current block is greater than a predetermined block size.

11 . The method of claim 10 , wherein the predetermined block size is 16.

12 . The method of claim 10 , wherein an initial motion vector of the current block is used as initial motion vector of the subblock.

13 . The method of claim 8 , wherein the second motion vector is obtained by using a pixel value indicated by a candidate motion vector in a region within an L0 reference picture for the current block, and using a pixel value indicated by a vector opposite to the candidate motion vector in a region within an L1 reference picture for the current block.

14 . The method of claim 8 , wherein the first motion vector of the current block is used for determining a boundary strength of deblocking filtering.

15 . An apparatus for transmitting data comprising a bitstream for an image, the apparatus comprising a processor configured to:
obtain the bitstream for the image, the bitstream for the image being generated by an image encoding method; and transmit the bitstream, wherein the image encoding method comprises: constructing a motion vector candidate list including a spatial motion vector and a temporal motion vector, from a neighboring block in a current picture including a current block and a col-block of the current block in a different picture from the current picture; determining a first motion vector of the current block using the motion vector candidate list; determining whether or not to perform motion vector refinement; obtaining a second motion vector of the current block by refining a first motion vector of the current block based on the determination of whether or not to perform motion vector refinement; generating a prediction block of the current block in an inter-prediction mode based on the second motion vector; generating a residual block of the current block based on the prediction block; generating transform coefficients of the current block based on the residual block of the current block; and encoding information on whether or not to perform motion vector refinement based on the determination of whether or not to perform motion vector refinement, wherein constructing the motion vector candidate list comprises:
entering a motion vector of the neighboring block as the spatial motion vector;
obtaining a refined motion vector of the col-block;
scaling th

Assignees

• Electronics & Telecommunications Res Inst
• Chips&Media Inc

Inventors

• Lim Sung Chang
• Kang Jung Won
• Lee Ha Hyun
• Lee Jin Ho
• Kim Hui Yong
• Kim Dae Yeon
• Park Dong Jin
• Jang Myoung Hun
• Jeong Wook Je

Classifications

• H04N19/52

  by predictive encoding · CPC title

• H04N19/184

  the unit being bits, e.g. of the compressed video stream · CPC title

• H04N19/176

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

• H04N19/119

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

• H04N19/139

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