Method of restoration in subblock units, and video decoding apparatus

A method for encoding a current block of video data, the method comprising:

• performing inter-prediction for the current block to generate predicted pixels of the current block;
• encoding, into a bitstream, transform coefficients corresponding to only part of sub-blocks in the current block, and sub-block information indicating a division form of the sub-blocks, wherein the transform coefficients are generated by transforming residual signals that are differences between original samples in the current block and the predicted samples;
• deriving a residual block of the current block from the transform coefficients on the basis of the sub-block information;
• and
• filtering boundaries of the sub-blocks in a reconstructed block of the current block derived on the basis of the residual block and the predicted samples,
• wherein the sub-block information includes:
• division direction information indicating a division direction of the sub-blocks;
• and
• at least one of division type information indicating a division type of the sub-blocks between a binary division type and a quad division type and position information indicating a position of at least one transform sub-block which is a transformed region among the sub-blocks,
• wherein the deriving of the residual block of the current block includes:
• identifying the transform sub-block and at least one non-transform sub-block which is a region not transformed, on the basis of the sub-block information;
• deriving residual samples corresponding to the transform sub-block, by inversely transforming the transform coefficients in the transform sub-block;
• and
• setting residual samples corresponding to the non-transform sub-block as zero.

A method for providing a video decoding apparatus with video data, the method comprising:

• encoding the video data into a bitstream;
• and
• transmitting the bitstream to the video decoding apparatus,
• wherein the encoding of the video data includes:
• performing inter-prediction for a current block of the video data to generate predicted pixels of the current block;
• encoding, into the bitstream, transform coefficients corresponding to only part of sub-blocks in the current block, and sub-block information indicating a division form of the sub-blocks, wherein the transform coefficients are generated by transforming residual signals that are differences between original samples in the current block and the predicted samples;
• deriving a residual block of the current block from the transform coefficients on the basis of the sub-block information;
• and
• filtering boundaries of the sub-blocks in a reconstructed block of the current block derived on the basis of the residual block and the predicted pixels,
• wherein the sub-block information includes:
• division direction information indicating a division direction of the sub-blocks;
• and
• at least one of division type information indicating a division type of the sub-blocks between a binary division type and a quad division type and position information indicating a position of at least one transform sub-block which is a transformed region among the sub-blocks,
• wherein the deriving of the residual block of the current block includes:
• identifying the transform sub-block and at least one non-transform sub-block which is a region not transformed, on the basis of the sub-block information;
• deriving residual samples corresponding to the transform sub-block, by inversely transforming the transform coefficients in the transform sub-block;
• and
• setting residual samples corresponding to the non-transform sub-block as zero.