Method and apparatus for removing redundancy in motion vector predictors
A method and apparatus derive a motion vector predictor (MVP) for a current block in an Inter, Merge, or Skip mode. Embodiments according to the present invention determine redundant MVP candidates according to a non-MV-value based criterion. The redundant MVP candidates are then removed from the MVP candidate set. In other embodiments according to the present invention, motion IDs are assigned to MVP candidates to follow the trail of motion vectors associated with the MVP candidate. An MVP candidate having a same motion ID as a previous MVP is redundant and can be removed from the MVP candidate set. In yet another embodiment, redundant MVP candidates correspond to one or more of the MVP candidates that cause the second 2N×N or N×2N PU to be merged into a 2N×2N PU are removed from the MVP candidate set.
1. A method of deriving a motion vector predictor (MVP) for a current block in an Inter, Merge, or Skip mode, the method comprising:
determining neighboring blocks of the current block, wherein an MVP candidate set is derived from MVP candidates associated with the neighboring blocks, wherein the neighboring blocks comprise a first N×N PU adjacent to an upper-left corner of the current block, a second N×N PU on the top of the current block, and a third N×N PU on the left of the current block, and wherein the current block corresponds to a fourth N×N PU (Prediction Unit) and;
determining at least one redundant MVP candidate, wherein an MVP candidate associated with the second N×N PU is determined to be said at least one redundant MVP candidate if the first N×N PU and the third N×N PU are merged such that the third N×N PU reuses motion information of the first N×N PU as an MVP, an MVP candidate associated with the third N×N PU is determined to be said at least one redundant MVP candidate if the first N×N PU and the second N×N PU are merged such that the second N×N PU reuses motion information of the first N×N PU as an MVP, and both the MVP candidates associated with the second and third N×N PUs are determined to be said at least one redundant MVP candidate if the first, second, and third N×N PUs are merged such that the third N×N PU and the second N×N both reuse motion information of the first N×N PU as respective MVPs;
removing said at least one redundant MVP candidate from the MVP candidate set; providing a modified MVP candidate set for determining a final MVP, wherein the modified MVP candidate set corresponds to the MVP candidate set with said at least one redundant MVP candidate removed; and
encoding or decoding the current block according to the final MVP.
2. The method of claim 1 , wherein an MVP candidate is determined to be said at least one redundant MVP candidate if said MVP candidate is within a same MDSR (Motion Data Storage Reduction) unit as another MVP candidate in the MVP candidate set.
3. The method of claim 2 , wherein said MVP candidate and said another MVP candidate use a same scaling factor.
4. The method of claim 3 , wherein said MVP candidate and said another MVP candidate use a same reference picture list.
5. The method of claim 1 , wherein an MVP candidate is determined to be said at least one redundant MVP candidate if said MVP candidate and another MVP candidate in the MVP candidate set are located in different PUs (Prediction Units) and the different PUs are merged.
6. The method of claim 1 , wherein the MVP candidate set comprises a first MVP candidate and a second MVP candidate, wherein one of said first MVP candidate and second MVP candidate is determined to be said at least one redundant MVP candidate if the first MVP candidate is coded as an Inter mode according to the second MVP candidate and zero MVD (Motion Vector Difference) and without scaling.
7. The method of claim 1 , wherein an MVP candidate is determined to be said at least one redundant MVP candidate if the current block corresponds to a second 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU (Prediction Unit) coded in a Merge mode and said MVP candidate causes the second 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU to be merged into a 2N×2N PU.
8. The method of claim 1 , wherein an MVP candidate is determined to be said at least one redundant MVP candidate if the current block corresponds to a second 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU (Prediction Unit) coded in a Merge mode and said MVP candidate located within a first 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU.
9. The method of claim 1 , wherein a syntax element is incorporated in a sequence, picture or slice header to indicate whether said at least one redundant MVP candidate is removed from the MVP candidate set.
10. An apparatus of deriving a motion vector predictor (MVP) for a current block in an Inter, Merge, or Skip mode, the apparatus comprising a circuit configured for:
determining neighboring blocks of the current block, wherein an MVP candidate set is derived from MVP candidates associated with the neighboring blocks, wherein the neighboring blocks comprise a first N×N PU adjacent to an upper-left corner of the current block, a second N×N PU on the top of the current block, and a third N×N PU on the left of the current block, and wherein the current block corresponds to a fourth N×N PU (Prediction Unit) and;
determining one or more redundant MVP candidates, wherein an MVP candidate associated with the second N×N PU is determined to be said at least one redundant MVP candidate if the first N×N PU and the third N×N PU are merged such that the third N×N PU reuses motion information of the first N×N PU as an MVP, an MVP candidate associated with the third N×N PU is determined to be said at least one redundant MVP candidate if the first N×N PU and the second N×N PU are merged such that the second N×N PU reuses motion information of the first N×N PU as an MVP, and both the MVP candidates associated with the second and third N×N PUs are determined to be said at least one redundant MVP candidate if the first, second, and third N×N PUs are merged such that the third N×N PU and the second N×N both reuse motion information of the first N×N PU as respective MVPs;
removing said one or more redundant MVP candidates from the MVP candidate set;
providing a modified MVP candidate set for determining a final MVP, wherein the modified MVP candidate set corresponds to the MVP candidate set with said one or more redundant MVP candidates removed; and
encoding or decoding the current block according to the final MVP.
11. The apparatus of claim 10 , wherein one or more of the MVP candidates are determined to be said one or more redundant MVP candidates if the current block corresponds to a second 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU (Prediction Unit) coded in a Merge mode and said one or more of the MVP candidates cause the second 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU to be merged into a 2N×2N PU.
12. The apparatus of claim 10 , wherein one or more of the MVP candidates are determined to be said one or more redundant MVP candidates if the current block corresponds to a second 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU (Prediction Unit) coded in a Merge mode and said one or more of the MVP candidates located within a first 2N×N, 2N×nU, 2N×nD, N×2N, nL×2N or nR×2N PU.
13. The apparatus of claim 10 , wherein a syntax element is incorporated in a sequence, picture or slice header to indicate whether said one or more redundant MVP candidates are removed from the MVP candidate set.