History-based image coding method and apparatus
An image decoding method performed by a decoding apparatus, includes deriving a history-based motion vector prediction (HMVP) buffer for a current block based on a history, and deriving motion information of the current block based on an HMVP candidate included in the HMVP buffer, thereby increasing inter prediction efficiency.
1. An image decoding method performed by a decoding apparatus, the method comprising:
deriving a history-based motion vector prediction (HMVP) buffer for a current block;
configuring motion information candidate list based on an HMVP candidate comprised in the HMVP buffer;
deriving motion information of the current block based on the motion information candidate list;
generating prediction samples for the current block based on the motion information; and
generating reconstructed samples based on the prediction samples,
wherein a current picture includes one or more tiles,
wherein the current block is for one of coding units (CUs) split from a coding tree unit (CTU),
wherein the HMVP buffer is updated based on motion information of a previous block,
wherein the HMVP buffer is initialized at a firstly-ordered CTU per each CTU row of each tile,
wherein the method further comprising:
determining whether the CTU is a firstly-ordered CTU in a CTU row of a current tile,
wherein the HMVP buffer for the current block is derived based on the result of the determination, and
wherein based on the result of the determining that the CTU is the firstly-ordered CTU in the CTU row of the current tile, the HMVP buffer is initialized for the CTU.
2. The method of claim 1 , wherein based on the HMVP buffer being initialized, the number of HMVP candidates comprised in the HMVP buffer is set to zero.
3. The method of claim 1 , wherein based on the current picture being split into a plurality of tiles, the HMVP buffer is initialized in units of CTU rows in each tile.
4. The method of claim 1 , wherein based on the result of the determination that the CTU comprising the current block is the firstly-ordered CTU of the CTU row in the current tile, the HMVP buffer comprises an HMVP candidate derived based on motion information of a block processed earlier than the current block in the CTU, and
wherein based on the result of the determination that the CTU is not the firstly-ordered CTU of the CTU row in the current tile, the HMVP buffer comprises an HMVP candidate derived based on motion information processed earlier than the current block in the CTU row in the current tile.
5. The method of claim 1 , wherein based on a merge mode or a motion vector prediction (MVP) mode being applied to the current block, the motion information candidate list is configured based on the HMVP candidate comprised in the HMVP buffer.
6. The method of claim 1 , wherein based on a merge mode being applied to the current block, the motion information candidate list is a merge candidate list, and the HMVP candidate comprised in the HMVP buffer is comprised as a candidate of the merge candidate list,
wherein the HMVP candidate among candidates comprised in the merge candidate list is indicated based on a merge index acquired from a bitstream.
7. An image encoding method performed by an encoding apparatus, the method comprising:
deriving a history-based motion vector prediction (HMVP) buffer for a current block;
configuring a motion information candidate list based on an HMVP candidate comprised in the HMVP buffer;
deriving motion information of the current block based on the motion information candidate list;
generating prediction samples for the current block based on the motion information;
deriving residual samples based on the prediction samples; and
encoding image information comprising information on the residual samples,
wherein one or more tiles are present in a current picture,
wherein the current block is for one of coding units (CUs) split from a coding tree unit (CTU),
wherein the HMVP buffer is updated based on motion information of a previous block,
wherein the HMVP buffer is initialized at a firstly-ordered CTU per each CTU row of each tile,
wherein the method further comprising:
determining whether the CTU is a firstly-ordered CTU in a CTU row of a current tile,
wherein the HMVP buffer for the current block is derived based on the result of the determination, and
wherein based on the result of the determining that the CTU is the firstly-ordered CTU in the CTU row of the current tile, the HMVP buffer is initialized for the CTU.
8. The method of claim 7 , wherein based on the HMVP buffer being initialized, the number of HMVP candidates comprised in the HMVP buffer is set to zero.
9. The method of claim 7 , wherein based on the current picture being split into a plurality of tiles, the HMVP buffer is initialized in units of CTU rows in each tile.
10. The method of claim 7 , wherein based on the result of the determining that the CTU comprising the current block is the firstly-ordered CTU of the CTU row in the current tile, the HMVP buffer comprises an HMVP candidate derived based on motion information of a block processed earlier than the current block in the CTU, and
wherein based on the result of the determining that the CTU is not the firstly-ordered CTU of the CTU row in the current tile, the HMVP buffer comprises an HMVP candidate derived based on motion information processed earlier than the current block in the CTU row in the current tile.
11. The method of claim 7 , wherein based on a merge mode or a motion vector prediction (MVP) mode being applied to the current block, the motion information candidate list is configured based on the HMVP candidate comprised in the HMVP buffer.
12. The method of claim 7 , wherein based on a merge mode being applied to the current block, the motion information candidate list is a merge candidate list, and the HMVP candidate comprised in the HMVP buffer is comprised as a candidate of the merge candidate list, and
wherein a merge index comprised in the image information indicates the HMVP candidate among candidates comprised in the merge candidate list.
13. A non-transitory computer-readable digital storage medium storing encoded information causing a decoding apparatus to perform an image decoding method, the method comprising:
deriving a history-based motion vector prediction (HMVP) buffer for a current block;
configuring motion information candidate list based on an HMVP candidate comprised in the HMVP buffer;
deriving motion information of the current block based on the motion information candidate list;
generating prediction samples for the current block based on the motion information; and
generating reconstructed samples based on the prediction samples,
wherein a current picture includes one or more tiles,
wherein the current block is for one of coding units (CUs) split from a coding tree unit (CTU),
wherein the HMVP buffer is updated based on motion information of a previous block,
wherein the HMVP buffer is initialized at a firstly-ordered CTU per each CTU row of each tile,
wherein the method further comprising:
determining whether the CTU is a firstly-ordered CTU in a CTU row of a current tile,
wherein the HMVP buffer for the current block is derived based on the result of the determination, and
wherein based on the result of the determining that the CTU is the firstly-ordered CTU in the CTU row of the current tile, the HMVP buffer is initialized for the CTU.