Coding method, device, system with merge mode
Methods, systems, and computer programs encoded on a computer storage medium for marking the availability of a potential candidate coding block for intra prediction mode estimation of a current coding block are provided. One example method includes obtaining a parent block of a current coding block (parentCurr), and a parent block of a candidate coding block (parentCand). The current coding block and the candidate coding block are each obtained by splitting of the parentCurr and the parentCand, respectively. In one aspect, the current coding block cannot use an intra prediction mode of a potential candidate coding block when parentCand and parentCurr are the same block, a function of a number of quadtree, ternary tree, and binary tree splitting operations is greater than a threshold minus 1, and a split mode of the parent block of the current coding block parentCurr is a ternary tree split.
1. A method of coding implemented by an encoding/decoding device, the method comprising:
obtaining a parent block of a current coding block (denoted as parentCurr) and a parent block of a candidate coding block (denoted as parentCand);
obtaining the current coding block by splitting of the parentCurr; and
obtaining the candidate coding block by splitting of the parentCand, wherein the current coding block cannot use an intra prediction mode of a potential candidate coding block, when:
the parentCand and the parentCurr are the same block,
a function of a number of quadtree, ternary tree, and binary tree splitting operations that are performed to obtain the candidate coding block is greater than a threshold minus 1, and
a split mode of the parentCurr is a ternary tree split.
2. The method according to claim 1 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the potential candidate coding block is marked unavailable.
3. The method according to claim 1 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the current coding block is predicted using intra prediction mode and when the potential candidate coding block is marked unavailable.
4. A method of coding implemented by an encoding/decoding device, the method comprising:
obtaining a parent block of a current coding block (denoted as parentCurr) and a parent block of a candidate coding block (denoted as parentCand);
obtaining the current coding block by splitting of the parentCurr; and
obtaining the candidate coding block by splitting of the parentCand, wherein the current coding block cannot use an intra prediction mode of a potential candidate coding block, when:
the parentCand and the parentCurr are the same block,
a function of a number of quadtree, ternary tree, and binary tree splitting operations that are performed to obtain the candidate coding block is greater than a threshold minus 1, and
a split mode of the parentCurr is a quadtree split and
both vertical coordinators of top-left samples of the current coding block and the candidate coding block are smaller than a vertical coordinator of a top-left sample of the same parent block plus half of a height of the parentCurr.
5. The method according to claim 4 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the potential candidate coding block is marked unavailable.
6. The method according to claim 4 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the current coding block is predicted using intra prediction mode and the potential candidate coding block is marked unavailable.
7. A method of coding implemented by an encoding/decoding device, the method comprising:
obtaining a parent block of a current coding block (denoted as parentCurr) and a parent block of a candidate coding block (denoted as parentCand);
obtaining the current coding block by splitting of the parentCurr; and
obtaining the candidate coding block by splitting of the parentCand, wherein the current coding block cannot use an intra prediction mode of a potential candidate coding block, when:
the parentCand and the parentCurr are the same block,
a function of a number of quadtree, ternary tree, and binary tree splitting operations that are performed to obtain the candidate coding block is greater than a threshold minus 1, and
a split mode of the parentCurr is a quadtree split, and
both vertical coordinators of top-left samples of the current coding block and the candidate coding block are greater than a vertical coordinator of a top-left sample of the same parent block plus half of a height of the parentCurr.
8. The method according to claim 7 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the potential candidate coding block is marked unavailable.
9. The method according to claim 7 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the current coding block is predicted using intra prediction mode and when the potential candidate coding block is marked unavailable.
10. A non-transitory computer program product comprising instructions which, when executed by one or more processors, cause the one or more processors to perform operations comprising:
obtaining a parent block of a current coding block (denoted as parentCurr) and a parent block of a candidate coding block (denoted as parentCand);
obtaining the current coding block by splitting of the parentCurr; and
obtaining the candidate coding block by splitting of the parentCand, wherein the current coding block cannot use an intra prediction mode of a potential candidate coding block, when:
the parentCand and the parentCurr are the same block,
a function of a number of quadtree, ternary tree, and binary tree splitting operations that are performed to obtain the candidate coding block is greater than a threshold minus 1, and
a split mode of the parentCurr is a ternary tree split.
11. The non-transitory computer program product according to claim 10 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the potential candidate coding block is marked unavailable.
12. The non-transitory computer program product according to claim 10 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the current coding block is predicted using intra prediction mode and when the potential candidate coding block is marked unavailable.
13. An encoder/decoder, comprising:
one or more processors; and
a non-transitory computer-readable storage medium coupled to the processors and storing instructions for execution by the processors, wherein the instructions, when executed by the one or more processors, configure the encoder/decoder to perform operations comprising:
obtaining a parent block of a current coding block (denoted as parentCurr) and a parent block of a candidate coding block (denoted as parentCand);
obtaining the current coding block by splitting of the parentCurr; and
obtaining the candidate coding block by splitting of the parentCand, wherein the current coding block cannot use an intra prediction mode of a potential candidate coding block, when:
the parentCand and the parentCurr are the same block,
a function of a number of quadtree, ternary tree, and binary tree splitting operations that are performed to obtain the candidate coding block is greater than a threshold minus 1, and
a split mode of the parentCurr is a ternary tree split.
14. The encoder/decoder according to claim 13 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the potential candidate coding block is marked unavailable.
15. The encoder/decoder according to claim 13 , wherein the intra prediction mode of the candidate coding block is not used in a prediction of the current coding block when the current coding block is predicted using intra prediction mode and when the potential candidate coding block is marked unavailable.