Inter-prediction method and apparatus in image coding system
An image decoding method performed by a decoding apparatus includes the operations of: deriving a correspondence block corresponding to the current block of the current picture, the correspondence block being located in a co-located picture; deriving a movement vector of the correspondence block; scaling the movement vector of the correspondence block based on a first temporal distance and second temporal distance, the first temporal distance being the difference between the POC values of the co-located picture and POC values of the reference picture of the correspondence block, and the second temporal distance being the difference between the POC values of the co-located picture and POC values of the current picture containing the current block; deriving the movement vector updated based on the scaled movement vector and location of the current block; and deriving a movement vector of the current block based on the updated movement vector.
1. A video decoding method performed by a video decoding apparatus, the method comprising:
obtaining a merge index information for a current block from a bitstream;
obtaining a first motion vector of a left block adjacent to a left side of the current block, wherein the current block is divided into a plurality of current sub-blocks;
deriving a plurality of collocated sub-blocks related to the current block in a current picture, wherein the plurality of collocated sub-blocks are located in a collocated picture and are derived based on the first motion vector of the left block;
deriving first sub-block motion vectors of the plurality of collocated sub-blocks in the collocated picture;
constructing a merge candidate list of the current block based on a merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks;
selecting the merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks from the merge candidate list based on the merge index information; and
deriving second sub-block motion vectors of the current sub-blocks based on the merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks,
wherein each of the second sub-block motion vectors is derived based on each of the first sub-block motion vectors.
2. The video decoding method of claim 1 , wherein a corresponding block located in the collocated picture is derived based on a right lower end corner neighboring block of a collocated block or a center right lower end block of the collocated block within the collocated picture, and
wherein the collocated block is a block at same position as that of the current block and having a size equal to that of the current block on the collocated picture.
3. The video decoding method of claim 1 , wherein the plurality of collocated sub-blocks for a corresponding block are in a specific area, and
wherein a size of the specific area is equal to a size of the current block.
4. The video decoding method of claim 1 , further comprising
deriving a corresponding block related to the current block in a current picture,
wherein the deriving the corresponding block comprises:
deriving a first block based on the first motion vector of the left block;
deriving a motion vector of the first block;
deriving an updated motion vector of the first block by scaling the motion vector of the first block based on a first temporal distance and a second temporal distance, wherein the first temporal distance is a difference between a picture order count (POC) value of the collocated picture and a POC value of a reference picture of the first block, the second temporal distance is a difference between the POC value of the collocated picture and a POC value of a current picture including the current block; and
deriving the corresponding block based on the updated motion vector.
5. The video decoding method of claim 4 , wherein the deriving the second sub-block motion vectors of the current sub-blocks comprises:
scaling the first sub-block motion vectors of the plurality of collocated sub-blocks in a specific area for the corresponding block based on the first temporal distance and the second temporal distance;
deriving a motion vector list by arranging scaled first sub-block motion vectors of the plurality of collocated sub-blocks in the specific area in order that a position pointed by each of the scaled first sub-block motion vectors is close to a specific reference point of the current block; and
deriving the second sub-block motion vectors of the current sub-blocks based on the motion vector list.
6. The video decoding method of claim 1 , wherein the merge candidate list includes a spatial candidate and a temporal candidate, and
wherein the merge candidate related to the first sub-block motion vectors of the plurality of collocated sub-blocks is included in the merge candidate list as additional temporal candidates.
7. The video decoding method of claim 1 , further comprising:
obtaining an index information for the current block from a bitstream;
constructing a motion vector predictor (MVP) candidate list of the current block based on a MVP candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks for the corresponding block; and
selecting the MVP candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks for the corresponding block based on the index information,
wherein the second sub-block motion vectors of the current sub-blocks are derived based on the selected MVP candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks for the corresponding block.
8. A video encoding method performed by a video encoding apparatus, the method comprising:
obtaining a first motion vector of a left block adjacent to a left side of a current block, wherein the current block is divided into a plurality of current sub-blocks;
deriving a plurality of collocated sub-blocks related to the current block in a current picture, wherein the plurality of collocated sub-blocks are located in a collocated picture and are derived based on the first motion vector of the left block;
deriving first sub-block motion vectors of the plurality of collocated sub-blocks in the collocated picture;
constructing a merge candidate list of the current block based on a merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks;
deriving second sub-block motion vectors of the current sub-blocks based on the merge candidate list;
generating a merge index information related to the merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks; and
encoding image information including the merge index information,
wherein each of the second sub-block motion vectors is derived based on each of the first sub-block motion vectors.
9. A non-transitory computer-readable digital storage medium storing a bitstream generated by a video encoding method, the method comprising:
obtaining a first motion vector of a left block adjacent to a left side of a current block, wherein the current block is divided into a plurality of current sub-blocks;
deriving a plurality of collocated sub-blocks related to the current block in a current picture, wherein the plurality of collocated sub-blocks are located in a collocated picture and are derived based on the first motion vector of the left block;
deriving first sub-block motion vectors of the plurality of collocated sub-blocks in the collocated picture;
constructing a merge candidate list of the current block based on a merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks;
deriving second sub-block motion vectors of the current sub-blocks based on the merge candidate list;
generating a merge index information related to the merge candidate including the first sub-block motion vectors of the plurality of collocated sub-blocks; and
encoding image information including the merge index information to output the bitstream,
wherein each of the second sub-block motion vectors is derived based on each of the first sub-block motion vectors.