Image decoding method and apparatus therefor

1 . A method for an image decoding, the method performed by a decoding apparatus and comprising:

selecting specific residual coefficients for sign prediction using a scan order for the sign prediction among the residual coefficients of a current block;

predicting signs for the specific residual coefficients;

deriving residual samples of the current block based on the predicted signs; and

generating a reconstructed picture based on the residual samples,

wherein the scan order for the sign prediction is derived based on an intra prediction mode applied to the current block,

wherein the selecting of the specific residual coefficients comprises calculating costs of the residual coefficients of the current block along the scan order, and selecting the specific residual coefficients for the sign prediction based on the costs of the residual coefficients, and

wherein the costs of the residual coefficients are derived based on values of the residual coefficient and values of a template, corresponding to positions of the residual coefficients, for border samples of the current block.

2 . The method of claim 1 , wherein the specific residual coefficients are M residual coefficients of the scan order.

3 . The method of claim 1 , wherein the scan order is derived as a vertical scan order based on the intra prediction mode of the current block being a horizontal intra prediction mode, and

wherein the scan order is derived as a horizontal scan order based on the intra prediction mode of the current block being a vertical intra prediction mode.

4 . The method of claim 1 , wherein the scan order is derived as a vertical scan order based on the intra prediction mode of the current block being one of specific horizontal directional intra prediction modes, the specific horizontal directional intra prediction modes are a HOR_IDX−offset intra prediction mode to a HOR_IDX+offset intra prediction mode, and a HOR_IDX intra prediction mode is a horizontal intra prediction mode, and

wherein the scan order is derived as a horizontal scan order based on the intra prediction mode of the current block being one of specific vertical directional intra prediction modes, the specific vertical directional intra prediction modes are a VER_IDX−offset intra prediction mode to a VER_IDX+offset intra prediction mode, and a VER_IDX intra prediction mode is a vertical intra prediction mode.

5 . The method of claim 1 , wherein the scan order is derived as a vertical scan order based on the intra prediction mode of the current block being a horizontal intra prediction mode, and

wherein the scan order is derived as a horizontal scan order based on the intra prediction mode of the current block being a vertical intra prediction mode.

6 . The method of claim 1 , wherein each of the costs of the residual coefficients of the current block is derived based on the following equation,

cost

=

❘

"\[LeftBracketingBar]"

C

i

,

j

❘

"\[RightBracketingBar]"

·

∑

k

=

0

N

-

1

❘

"\[LeftBracketingBar]"

T

i

,

j

(

k

)

❘

"\[RightBracketingBar]"

where, cost represents the cost of the residual coefficient, C i,j is a value of the residual coefficient at position (i, j), T i,j is a template corresponding to position (i, j), and N represents a number of border pixels of the current block.

7 . The method of claim 1 , wherein deriving the residual samples of the current block based on the predicted signs includes:

determining whether the predicted signs are accurate based on syntax elements for the predicted signs; and

deriving signs of the specific residual coefficients based on a result of the determination,

wherein a target syntax element among the syntax elements represents whether a predicted sign for the target syntax element is accurate.

8 . The method of claim 7 , wherein the number of context-coded bins for the current block, including the context-coded bins of the syntax elements for the predicted signs, is limited to a specific threshold.

9 . The method of claim 7 , wherein context coded bins of residual information of the current block excluding the syntax elements are limited to a first specific number, and

wherein context coded bins of the syntax elements are limited to a second specific number.

10 . A method for an image encoding, the method performed by an encoding apparatus and comprising:

selecting specific residual coefficients for sign prediction using a scan order for the sign prediction among residual coefficients of a current block;

predicting signs for the specific residual coefficients; and

encoding residual information of the current block including syntax elements for the predicted signs,

wherein a target syntax element among the syntax elements represents whether the predicted sign for the target syntax element is accurate, and

wherein the scan order for the sign prediction is derived based on an intra prediction mode applied to the current block,

wherein the selecting of the specific residual coefficients comprises calculating costs of the residual coefficients of the current block along the scan order, and selecting the specific residual coefficients for the sign prediction based on the costs of the residual coefficients, and

wherein the costs of the residual coefficients are derived based on values of the residual coefficient and values of a template, corresponding to positions of the residual coefficients, for border samples of the current block.

11 . The method of claim 10 , wherein the specific residual coefficients are M residual coefficients of the scan order.

12 . The method of claim 10 , wherein the scan order is derived as a vertical scan order based on the intra prediction mode of the current block being a horizontal intra prediction mode, and

wherein the scan order is derived as a horizontal scan order based on the intra prediction mode of the current block being a vertical intra prediction mode.

13 . The method of claim 10 , wherein the scan order is derived as a vertical scan order based on the intra prediction mode of the current block being one of specific horizontal directional intra prediction modes, the specific horizontal directional intra prediction modes are a HOR−HOR_OFFSET intra prediction mode to a HOR+HOR_OFFSET intra prediction mode, and a HOR intra prediction mode is a horizontal intra prediction mode, and

wherein the scan order is derived as a horizontal scan order based on the intra prediction mode of the current block being one of specific vertical directional intra prediction modes, the specific vertical directional intra prediction modes are a VER−VER_OFFSET intra prediction mode to a VER+VER_OFFSET intra prediction mode, and a VER intra prediction mode is a vertical intra prediction mode.

14 . A method for transmitting data for image, the method comprising:

obtaining a bitstream of image information including residual information including syntax elements for predicted signs of specific residual coefficients of a current block; and

transmitting the data including the bitstream of the image information including the residual information,

wherein a target syntax element among the syntax elements represents whether a predicted sign for the target syntax element is accurate,

wherein the specific residual coefficients for sign prediction are selected as a scan order among residual coefficients of the current block, and

wherein the scan order for the sign prediction is derived based on an intra prediction mode applied to the current block,

wherein the selecting of the specific residual coefficients comprises calculating costs of the residual coefficients of the current block along the scan order, and selecting the specific residual coefficients for the sign prediction based on the costs of the residual coefficients, and

wherein the costs of the residual coefficients are derived based on values of the residual coefficient and values of a template, corresponding to positions of the residual coefficients, for border samples of the current block.