Method and apparatus for processing video signal by applying secondary transform to partitioned block

1. A method for reconstructing a video signal, the method comprising:

obtaining a transform block by performing an entropy decoding and a dequantization for a current block;

deriving a secondary transform matrix corresponding to a specific area in the transform block, wherein the specific area represents an area including a top-left area of the transform block and wherein the secondary transform matrix is denoted based on an input length and an output length of a secondary transform;

performing an inverse secondary transform for the specific area based on the secondary transform matrix;

performing an inverse primary transform for a block which the inverse secondary transform is applied to; and

reconstructing the current block based on a block which the primary inverse transform is applied to,

wherein the input length of the inverse secondary transform is determined as 16, and the output length of the inverse secondary transform is determined as 48, based on that each of the height and the width of the transform block is equal to 8 or 16.

2. The method of claim 1 , wherein the specific area is divided into 4×4 subblocks, and

wherein the inverse secondary transform is performed for each of the 4×4 subblocks.

3. The method of claim 2 , wherein the same 4×4 secondary transform is applied or different 4×4 secondary transforms are applied to the 4×4 subblocks based on at least one of locations or prediction modes of the subblocks.

4. The method of claim 2 , wherein whether the specific area is divided into 4×4 subblocks is determined based on a size of the transform block.

5. The method of claim 2 , further comprising

checking whether the number of non-zero transform coefficients within in the 4×4 subblocks is equal to or more than a specific threshold,

wherein whether the 4×4 secondary transform is applied to the 4×4 subblocks is determined according to the checking result.

6. The method of claim 5 , wherein when the number of non-zero transform coefficients within the 4×4 subblocks is equal to or more than the specific threshold, the 4×4 secondary transform is applied to the 4×4 subblocks, and

otherwise, the 4×4 secondary transform is not applied to the 4×4 subblocks.

7. An apparatus for reconstructing a video signal, the apparatus comprising:

a processor configured to

perform an entropy decoding for a current block;

perform a dequantization for the current block to obtain a transform block;

derive a secondary transform matrix corresponding to a specific area in the transform block, wherein the specific area represents an area including a top-left area of the transform block and wherein the secondary transform matrix is denoted based on an input length and an output length of a secondary transform,

perform an inverse secondary transform for the specific area based on the secondary transform matrix;

perform an inverse primary transform for a block which the inverse secondary transform is applied; and

reconstruct the current block based on a block which the inverse primary transform is applied to,

wherein the input length of the inverse secondary transform is determined as 16, and the output length of the inverse secondary transform is determined as 48, based on that each of the height and the width of the transform block is equal to 8 or 16.

8. The apparatus of claim 7 , wherein the specific area is divided into 4×4 subblocks, and

wherein the inverse secondary transform is performed for each of the 4×4 subblocks.

9. The apparatus of claim 8 , wherein the same 4×4 secondary transform is applied or different 4×4 secondary transforms are applied to the 4×4 subblocks based on at least one of locations or prediction modes of the subblocks.

10. The apparatus of claim 8 , wherein whether the specific area is divided into 4×4 subblocks is determined based on a size of the transform block.

11. The apparatus of claim 8 , wherein the processor is further configured to:

check whether the number of non-zero transform coefficients within the 4×4 subblocks is equal to or more than a specific threshold and determines whether the 4×4 secondary transform is applied to the 4×4 subblocks according to the checking result.

12. The apparatus of claim 11 , wherein when the number of non-zero transform coefficients within the 4×4 subblocks is equal to or more than the specific threshold, the 4×4 secondary transform is applied to the 4×4 subblocks, and

otherwise, the 4×4 secondary transform is not applied to the 4×4 subblocks.

13. A non-transitory computer-readable medium storing a bitstream causing an image decoding apparatus to perform an image decoding method, the image decoding method comprising:

obtaining a transform block by performing an entropy decoding and a dequantization for a current block;

deriving a secondary transform matrix corresponding to a specific area in the transform block, wherein the specific area represents an area including a top-left area of the transform block and wherein the secondary transform matrix is denoted based on an input length and an output length of a secondary transform;

performing an inverse secondary transform for the specific area based on the secondary transform matrix;

performing an inverse primary transform for a block which the inverse secondary transform is applied to; and

reconstructing the current block based on a block which the primary inverse transform is applied to,

wherein the input length of the inverse secondary transform is determined as 16, and the output length of the inverse secondary transform is determined as 48, based on that each of the height and the width of the transform block is equal to 8 or 16.