Method and apparatus for decoding image by using transform according to block size in image coding system

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

obtaining a flag for whether transform coefficients for a current block are present;

deriving prediction samples by performing a prediction for the current block;

determining whether the current block is partitioned based on the flag, a size of the current block and a maximum transform size;

deriving transform blocks by partitioning the current block based on a result of the determination;

deriving the transform coefficients for the derived transform blocks;

deriving residual samples by performing an inverse transform process for the transform coefficients; and

generating a reconstructed picture based on the residual samples and the prediction samples,

wherein based on the flag informing that the transform coefficients are not present, it is determined that the current block is not partitioned into the transform blocks, and

wherein based on the flag informing the transform coefficients are present, it is determined that the current block is partitioned into the transform blocks based on at least one of a case in which a width of the current block is greater than a width of the maximum transform size and a case a height of the current block is greater than a height of the maximum transform size,

wherein the width of the current block is W and the height of the current block is H,

wherein based on the width of the current block being greater than the width of the maximum transform size, a width of the transform blocks is derived as W/2, and

wherein based on the height of the current block being greater than the height of the maximum transform size, a height of the transform blocks is derived as H/2.

2. The method of claim 1 , wherein based on the width of the current block being greater than the maximum transform size and the height of the current block being not greater than the height of the maximum transform size, the current block is partitioned into the transform blocks of (W/2)×H size.

3. The method of claim 1 , wherein based on the width of the current block being not greater than the width of the maximum transform size and the height of the current block being greater than the height of the maximum transform size, the current block is partitioned into the transform blocks of W×(H/2) size.

4. The method of claim 1 , wherein based on the width of the current block being greater than the width of the maximum transform size and the height of the current block being greater than the height of the maximum transform size, the current block is partitioned into the transform blocks of (W/2)×(H/2) size.

5. The method of claim 1 , wherein based on the width of the current block being greater than the width of maximum transform size and the height of the current block being not greater than the height of the maximum transform size, the current block is partitioned into the transform blocks of H×H size.

6. The method of claim 1 , wherein based on the current block being a non-square block, a width of the current block being not larger than the maximum transform size and a height of the current block being larger than the maximum transform size, the current block is partitioned into the transform blocks of W×W size.

7. The method of claim 1 , wherein the size of the current block is 128×128, and the maximum transform size is 64×64.

8. The method of claim 1 , wherein the maximum transform size is a size of a maximum transform kernel among transform kernels.

9. The method of claim 1 , further comprising performing deblocking filtering for a reconstructed sample located in an adjacent area to boundaries of the transform blocks in the reconstructed picture.

10. A method of encoding video performed by an encoding device, the method comprising:

deriving prediction samples by performing a prediction for a current block;

deriving residual samples based on original samples and the prediction samples for the current block;

deriving a flag for whether transform coefficients for the current block are present;

determining whether the current block is partitioned based on the flag, a size of the current block and a maximum transform size;

deriving transform blocks by partitioning the current block based on a result of the determination;

deriving the transform coefficients by performing a transform process on residual samples included in the derived transform blocks;

deriving quantized transform coefficients by performing a quantization process on the transform coefficients; and

encoding video information comprising the flag and information for the quantized transform coefficients,

wherein based on the flag informing that the transform coefficients are not present, it is determined that the current block is not partitioned into the transform blocks, and

wherein based on the flag informing the transform coefficients are present, it is determined that the current block is partitioned into the transform blocks based on at least one of a case in which a width of the current block is greater than a width of the maximum transform size and a case a height of the current block is greater than a height of the maximum transform size,

wherein the width of the current block is W and the height of the current block is H,

wherein based on the width of the current block being greater than the width of the maximum transform size, a width of the transform blocks is derived as W/2, and

wherein based on the height of the current block being greater than the height of the maximum transform size, a height of the transform blocks is derived as H/2.

11. A non-transitory computer-readable storage medium storing a bitstream, the bitstream, when executed, causing a decoding apparatus to perform the following steps:

obtaining a flag for whether transform coefficients for a current block are present;

deriving prediction samples by performing a prediction for the current block;

determining whether the current block is partitioned based on the flag, a size of the current block and a maximum transform size;

deriving transform blocks by partitioning the current block based on a result of the determination;

deriving the transform coefficients for the derived transform blocks;

deriving residual samples by performing an inverse transform process for the transform coefficients; and

generating a reconstructed picture based on the residual samples and the prediction samples,

wherein based on the flag informing that the transform coefficients are not present, it is determined that the current block is not partitioned into the transform blocks, and

wherein based on the flag informing the transform coefficients are present, it is determined that the current block is partitioned into the transform blocks based on at least one of a case in which a width of the current block is greater than a width of the maximum transform size and a case a height of the current block is greater than a height of the maximum transform size,

wherein the width of the current block is W and the height of the current block is H,

wherein based on the width of the current block being greater than the width of the maximum transform size, a width of the transform blocks is derived as W/2, and

wherein based on the height of the current block being greater than the height of the maximum transform size, a height of the transform blocks is derived as H/2.