Method and device for processing video signal by using affine motion prediction
Disclosed are a method for processing a video signal by using an affine motion prediction and an apparatus thereof. A method for processing a video signal according to the present disclosure may include: adding, to an affine candidate list, an affine coded block coded in an affine prediction mode among neighbor blocks of a current block; obtaining a syntax element indicating a candidate used for the affine motion prediction of the current block in the affine candidate list; deriving a control point motion vector predictor of the current block based on an affine motion model of the candidate indicated by the syntax element; deriving a control point motion vector of the current block by adding a control point motion vector difference to the control point motion vector predictor; and generating a prediction block of the current block by using the control point motion vector of the current block.
1. A method for decoding a video signal, comprising:
determining a block coded in an affine prediction mode among neighbor blocks of a current block as an affine candidate based on that the block having a same reference picture as a reference picture of the current block;
obtaining a syntax element related to an affine candidate used for the affine motion prediction of the current block among affine candidates;
deriving a control point motion vector predictor of the current block based on the affine candidate related to the syntax element;
deriving a control point motion vector of the current block based on a control point motion vector difference and the control point motion vector predictor;
generating prediction samples of the current block based on the control point motion vector of the current block; and
generating reconstructed samples of the current block based on the prediction samples of the current block,
wherein the determining of the block further includes:
grouping the neighbor blocks of the current block into a first group and a second group; and
searching the block coded in the affine prediction mode based on a predefined order in each of the first group and the second group.
2. The method of claim 1 , wherein the determining of the block determines a predefined maximum number of blocks coded in the affine prediction mode as the affine candidates, and
wherein the predefined maximum number is two.
3. The method of claim 1 , wherein the determining of the block comprises:
searching a valid affine coded block among the blocks having the same reference picture as the reference picture of the current block in the first group according to a predefined order, and
searching a valid affine coded block among the blocks having the same reference picture as the reference picture of the current block in the top group according to a predefined order.
4. The method of claim 1 , wherein the determining of the block further includes: determining an affine candidate scaled based on a picture order count between reference pictures when there is no affine coded block among the neighbor blocks of the current block.
5. The method of claim 1 , wherein the determining of the block further includes:
determining 0 or 1 block coded in the affine prediction mode in each group as the affine candidate.
6. A method for encoding a video signal, comprising:
determining a block coded in an affine prediction mode among neighbor blocks of a current block as an affine candidate based on that the block having a same reference picture as a reference picture of the current block;
selecting an affine candidate used for the affine motion prediction of the current block among affine candidates;
deriving a control point motion vector predictor of the current block based on the selected affine candidate;
deriving a control point motion vector of the current block based on a control point motion vector difference and the control point motion vector predictor;
generating prediction samples of the current block based on the control point motion vector of the current block;
generating residual samples of the current block based on the prediction samples; and
generating a syntax element related to the selected affine candidate among the affine candidates,
wherein the determining of the block further includes:
grouping the neighbor blocks of the current block into a first group and a second group; and
searching the block coded in the affine prediction mode based on a predefined order in each of the first group and the second group.
7. The method of claim 6 , wherein the determining of the block determines a predefined maximum number of blocks coded in the affine prediction mode as the affine candidates, and
wherein the predefined maximum number is two.
8. The method of claim 6 , wherein the determining of the block comprises:
searching a valid affine coded block among the blocks having the same reference picture as the reference picture of the current block in the first group according to a predefined order, and
searching a valid affine coded block among the blocks having the same reference picture as the reference picture of the current block in the top group according to a predefined order.
9. The method of claim 6 , wherein the determining of the block further includes:
determining 0 or 1 block coded in the affine prediction mode in each group as the affine candidate.
10. A non-transitory decoder-readable storage medium storing encoded picture information generated by performing the steps of:
determining a block coded in an affine prediction mode among neighbor blocks of a current block as an affine candidate based on that the block having a same reference picture as a reference picture of the current block;
selecting an affine candidate used for the affine motion prediction of the current block among affine candidates;
deriving a control point motion vector predictor of the current block based on the selected affine candidate;
deriving a control point motion vector of the current block based on a control point motion vector difference and the control point motion vector predictor;
generating prediction samples of the current block based on the control point motion vector of the current block;
generating residual samples of the current block based on the prediction samples; and
generating a syntax element related to the selected affine candidate among the affine candidates,
wherein the determining of the block further includes:
grouping the neighbor blocks of the current block into a first group and a second group; and
searching the block coded in the affine prediction mode based on a predefined order in each of the first group and the second group.