Method and device for video coding using various transform techniques

1. A method of decoding a video, the method comprising:

determining, based on a sub-block transform flag, whether a sub-block transform is performed on a current coding block, wherein the sub-block transform flag indicates whether the sub-block transform is used for the current coding block, and the sub-block transform refers to a transform method in which inverse transform is performed on a first sub-block of two sub-blocks in the current coding block and the inverse transform is not performed on a second sub-block of the two sub-blocks;

determining, based on the sub-block transform flag indicating that the sub-block transform is used for the current coding block, a split type for the sub-block transform of the current coding block among a plurality of split type candidates including a symmetric binary split and an asymmetric binary split;

splitting the current coding block into the first sub-block and the second sub-block based on the split type for the sub-block transform;

determining horizontal and vertical transform types of the first sub-block; and

performing the inverse transform on the first sub-block based on the horizontal and vertical transform types,

wherein one of the two sub-blocks is determined as the first sub-block based on position information indicating a position of a sub-block on which the inverse transform is performed among the two sub-blocks,

wherein a left or top sub-block of the two sub-blocks is determined as the first sub-block based on a value of the position information being equal to a first value, and a right or bottom sub-block of the two sub-blocks is determined as the first sub-block based on the value of the position information being equal to a second value, and

wherein the horizontal and vertical transform types are determined based on the value of the position information.

2. The method of claim 1 , wherein the method further comprises determining a split direction of the current coding block based on the sub-block transform flag indicating that the sub-block transform is performed on the current coding block.

3. The method of claim 1 , wherein, based on the first sub-block being the top sub-block of the two sub-blocks in the current coding block, the horizontal transform type is different from the vertical transform type, and

wherein, based on the first sub-block being the bottom sub-block of the two sub-blocks in the current coding block, the horizontal transform type is the same as the vertical transform type.

4. The method of claim 1 , wherein, based on the first sub-block being the left sub-block of the two sub-blocks in the current coding block, the horizontal transform type is different from the vertical transform type, and

wherein, based on the first sub-block being the right sub-block of the two sub-blocks in the current coding block, the horizontal transform type is the same as the vertical transform type.

5. The method of claim 1 , wherein the sub-block transform flag is obtained from a bitstream based on prediction mode of the current coding block being an inter prediction mode.

6. A method of encoding a video, the method comprising:

determining whether a sub-block transform is performed on a current coding block, wherein the sub-block transform refers to a transform method in which transform is performed on a first sub-block of two sub-blocks in the current coding block and the transform is not performed on a second sub-block of the two sub-blocks;

determining, based on the determination that the sub-block transform is performed on the current coding block, a split type for the sub-block transform of the current coding block among a plurality of split type candidates including a symmetric binary split and an asymmetric binary split;

determining horizontal and vertical transform types of the first sub-block resulting from splitting the current coding block based on the split type; and

performing the transform on the first sub-block based on the horizontal and vertical transform types,

wherein position information indicating a position of the first sub-block on which the transform is performed among the two sub-blocks is encoded into a bitstream,

wherein the position information is encoded with a first value based on a left or top sub-block of the two sub-blocks being the first sub-block, and the position information is encoded with a second value based on a right or bottom sub-block of the two sub-blocks being the first sub-block, and

wherein the horizontal and vertical transform types are determined based on the value of the position information.

7. A method for transmitting a bitstream associated with a video signal, comprising:

determining whether a sub-block transform is performed on a current coding block, wherein the sub-block transform refers to a transform method in which transform is performed on a first sub-block of two sub-blocks in the current coding block and the transform is not performed on a second sub-block of the two sub-blocks;

determining, based on the determination that the sub-block transform is performed on the current coding block, a split type for the sub-block transform of the current coding block among a plurality of split type candidates including a symmetric binary split and an asymmetric binary split;

determining horizontal and vertical transform types of the first sub-block resulting from splitting the current coding block based on the split type;

performing the transform on the first sub-block based on the horizontal and vertical transform types to obtain transform coefficients of the first sub-block;

encoding the transform coefficients of the first sub-block to generate the bitstream; and

transmitting the bitstream,

wherein position information indicating a position of the first sub-block on which the transform is performed among the two sub-blocks is encoded into the bitstream,

wherein the position information is encoded with a first value based on a left or top sub-block of the two sub-blocks being the first sub-block, and the position information is encoded with a second value based on a right or bottom sub-block of the two sub-blocks being the first sub-block, and

wherein the horizontal and vertical transform types are determined based on the value of the position information.