Image coding method based on secondary transform, and device therefor
An image decoding method according to the present document comprises the steps of: deriving transform coefficients through inverse quantization on the basis of quantized transform coefficients for a target block; deriving modified transform coefficients on the basis of an inverse reduced secondary transform (RST) for the transform coefficients; and generating, on the basis of an inverse primary transform for the modified transform coefficients, a restoration picture based on residual samples for the target block, wherein the modified transform coefficients derived according to the inverse RST are two-dimensionally arranged according to the order of a row priority direction or a column priority direction according to an intra prediction mode to be applied to the target block.
1. An image decoding method performed by a decoding apparatus, the method comprising:
deriving transform coefficients for a target block from residual information;
deriving modified transform coefficients based on an inverse non-separable secondary transform for the transform coefficients;
deriving residual samples for the target block based on an inverse primary transform for the modified transform coefficients; and
generating a reconstructed picture based on the residual samples for the target block,
wherein the inverse non-separable secondary transform derives a larger number of output data than input data through a matrix operation of the input data for a top-left area of the target block and a transform kernel matrix, and
wherein the modified transform coefficients are two-dimensionally arranged according to a row-first direction, by applying the inverse non-separable secondary transform, based on an intra prediction mode applied to the target block being a horizontal mode in a left direction with respect to a top-left diagonal direction.
2. The image decoding method of claim 1 , wherein, when an intra prediction mode applicable to the target block is one of 65 directional modes, an intra prediction mode in the top-left diagonal direction is mode 34, and the intra prediction mode applied to the target block comprises mode 2 to mode 34, the modified transform coefficients are two-dimensionally arranged according to the row-first direction.
3. The image decoding method of claim 1 , wherein, when the intra prediction mode applied to the target block is a planar mode or a DC mode, the modified transform coefficients are two-dimensionally arranged according to the row-first direction.
4. The image decoding method of claim 1 , wherein the inverse non-separable secondary transform is performed based on a transform set determined based on a mapping relationship according to the intra prediction mode applied to the target block and the transform kernel matrix selected from among a plurality of transform kernel matrices comprised in each of the transform set, and is performed based on a transform index related to whether the inverse non-separable secondary transform is applied and one of the transform kernel matrix comprised in the transform set.
5. The image decoding method of claim 1 , wherein the input data for the top-left area is in a top-left 4×4 area of the target block and the output data for the top-left area is in the top-left 4×4 area of the target block or a top-left 8×8 area of the target block.
6. An image encoding method performed by an image encoding apparatus, the method comprising:
deriving residual samples for a target block;
deriving transform coefficients for the target block based on a primary transform for the residual samples;
deriving modified transform coefficients based on a non-separable secondary transform for the transform coefficients; and
encoding residual information based on the modified transform coefficients,
wherein the non-separable secondary transform derives a smaller number of output data than input data through a matrix operation of the input data for a top-left area of the target block and a transform kernel matrix,
wherein the transform coefficients are one-dimensionally arranged according to a row-first direction for applying the non-separable secondary transform, based on an intra prediction mode applied to the target block being a horizontal mode in a left direction with respect to a top-left diagonal direction.
7. The image encoding method of claim 6 , wherein, when an intra prediction mode applicable to the target block is one of 65 directional modes, an intra prediction mode in the top-left diagonal direction is mode 34, and the intra prediction mode applied to the target block comprises mode 2 to mode 34, the transform coefficients are one-dimensionally arranged according to the row-first direction.
8. The image encoding method of claim 6 , wherein, when the intra prediction mode applied to the target block is a planar mode or a DC mode, the 8 transform coefficients are one-dimensionally arranged according to the row-first direction.
9. The image encoding method of claim 6 , wherein the input data for the top-left area is in a top-left 4×4 area of the target block or a top-left 8×8 area of the target block and the output data for the top-left area is in the top-left 4×4 area of the target block.
10. A non-transitory computer readable storage medium storing a bitstream generated by a method, the method comprising:
deriving residual samples for a target block;
deriving transform coefficients for the target block based on a primary transform for the residual samples;
deriving modified transform coefficients based on a non-separable secondary transform for the transform coefficients; and
encoding residual information based on the modified transform coefficients to generate the bitstream,
wherein the non-separable secondary transform derives a smaller number of output data than input data through a matrix operation of the input data for a top-left area of the target block and a transform kernel matrix,
wherein the transform coefficients are one-dimensionally arranged according to a row-first direction for applying the non-separable secondary transform, based on an intra prediction mode applied to the target block being a horizontal mode in a left direction with respect to a top-left diagonal direction.