Method and apparatus for deriving temporal motion vector prediction

Patent metadata

Field	Value
Publication number	US-10412406-B2
Application number	US-201615564206-A
Country	US
Kind code	B2
Filing date	Apr 13, 2016
Priority date	Apr 14, 2015
Publication date	Sep 10, 2019
Grant date	Sep 10, 2019

Abstract

A method and apparatus for encoding or decoding a motion vector (MV) of a current block of a current picture using advanced temporal motion vector prediction are disclosed. At least two collocated pictures are selected from a set of reference pictures of the current picture. One or more TMVPs are derived based on reference motion vectors (MVs) associated with collocated reference blocks of the collocated pictures. A motion vector prediction candidate set including one or more TMVPs is then determined. The current block is encoded or decoding using the motion vector prediction candidate set. The reference motion vectors (MVs) are scaled before the reference motion vectors (MVs) are used to derive the TMVPs.

First claim

The invention claimed is:

1. A method of encoding or decoding a motion vector (MV) of a current block of a current picture using motion vector prediction, comprising:
selecting a plurality of collocated pictures from a set of reference pictures of the current picture for the current block;
identifying a plurality of collocated reference blocks from the plurality of collocated pictures;
obtaining a plurality of reference motion vectors (MVs) associated with the plurality of collocated reference blocks, wherein each reference MV is associated with a specific collocated reference block in a specific collocated picture among the plurality of collocated pictures and at least one collocated reference picture of the specific collocated picture;
deriving one or more temporal motion vector predictors (MVPs) based on the plurality of reference motion vectors (MVs);
determining a motion vector prediction candidate set including said one or more TMVPs; and
encoding or decoding the current block using the motion vector prediction candidate set.

2. The method of claim 1 , wherein said reference motion vectors (MVs) are scaled before said reference motion vectors (MVs) are used to derive said one or more TMVPs.

3. The method of claim 1 , wherein a total picture number corresponding to said at least two collocated pictures is fixed to a predetermined value or wherein a total picture number corresponding to said at least two collocated pictures is determined by an encoder.

4. The method of claim 3 , wherein the total picture number corresponding to said at least two collocated pictures is signalled from the encoder to a decoder in video parameter set (VPS), sequence parameter set (SPS), picture parameter set (PPS), slice header, coding unit (CU), or prediction unit (PU).

5. The method of claim 1 , wherein said at least two collocated pictures are located at fixed positions in a reference picture list.

6. The method of claim 1 , wherein said at least two collocated pictures are selected by an encoder and positions of said at least two collocated pictures in a reference picture list are signalled from the encoder to a decoder.

7. The method of claim 6 , wherein the positions of said at least two collocated pictures in a reference picture list are signalled from the encoder to a decoder in video parameter set (VPS), sequence parameter set (SPS), picture parameter set (PPS), slice header, coding unit (CU), or prediction unit (PU).

8. The method of claim 1 , wherein one reference motion vector associated with a collocated reference block for each collocated picture is selected for said deriving said one or more TMVPs.

9. The method of claim 8 , wherein an encoder selects one TMVP associated with the selected reference motion vector and the encoder signals selection information in a bitstream.

10. The method of claim 1 , wherein said deriving one or more TMVPs selects N TMVPs from different collocated reference blocks in different collocated pictures and said N TMVPs are inserted into the motion vector prediction candidate set corresponding to a merging candidate list or an advanced motion vector prediction (AMVP) candidate list, and wherein N is a positive integer.

11. The method of claim 10 , wherein said deriving one or more TMVPs selects said N TMVPs from different collocated reference blocks in different collocated pictures according to a selected checking order.

12. The method of claim 10 , wherein when N is equal to one, said deriving one or more TMVPs checks one or more collocated blocks in a first collocated picture first; if any reference motion vector associated with said one or more collocated blocks is found, said deriving one or more TMVPs selects the reference motion vector found as one TMVP; and otherwise said deriving one or more TMVPs selects one TMVP from a next collocated picture.

13. The method of claim 10 , wherein when N is equal to two, said deriving one or more TMVPs checks one or more collocated blocks of a first collocated picture in a selected checking order; if a first motion vector that is a firstly available reference motion vector associated with one checked collocated block of the first collocated picture in the selected checking order is found, the first motion vector is selected as a first TMVP and checking remaining collocated blocks of the first collocated picture is skipped; said deriving one or more TMVPs then checks one or more collocated blocks of a second collocated picture in the selected checking order; and if a second motion vector that is a firstly available reference motion vector associated with one checked collocated block of the second collocated picture in the selected checking order is found, the second motion vector is selected as a second TMVP and checking remaining collocated blocks of the second collocated picture is skipped.

14. The method of claim 1 , wherein said reference motion vectors (MVs) associated with the collocated reference blocks of said at least two collocated pictures are allowed to be stored in a spatial compressed format, and different collocated pictures are allowed to use different compression ratios.

15. The method of claim 14 , wherein an encoder transmits one or more syntax elements to specify whether said reference motion vectors (MVs) associated with the collocated reference blocks of said at least two collocated pictures are stored in the spatial compressed format and said one or more syntax elements are signalled in a video parameter set (VPS), sequence parameter set (SPS), picture parameter set (PPS), or slice header.

16. The method of claim 1 , wherein a first TMVP and a second TMVP are combined to form a bi-prediction TMVP with a first motion vector in a first reference list and a second motion vector in a second reference list, and wherein the first motion vector is obtained by scaling the first TMVP and the second motion vector is obtained by scaling the second TMVP.

17. The method of claim 1 , wherein the reference motion vectors (MVs) associated with the collocated reference blocks of said at least two collocated pictures are scaled to a fixed time-distance before the reference motion vectors (MVs) are stored in a buffer.

18. The method of claim 17 , wherein reference picture indices for the reference motion vectors (MVs) are not stored.

19. The method of claim 1 , further comprising determining a number of remaining candidates of the motion vector prediction candidate set, if the number of remaining candidates is less than or equal to a threshold, only one TMVP is included in the motion vector prediction candidate set, otherwise two or more TMVPs are included in the motion vector prediction candidate set.

20. An apparatus for encoding or decoding a motion vector (MV) of a current block of a current picture using motion vector prediction, comprising one or more electronic circuits arranged to:
select a plurality of collocated pictures from a set of reference pictures of the current picture for the current block;
identify a plurality of collocated reference blocks from the plurality of collocated pictures;
obtain a plurality of reference motion vectors (MVs) associated with the plurality of collocated reference blocks, wherein each reference MV is associated with a specific collocated reference block in a specific collocated picture among the plurality of collocated pictures and at least one collocated reference picture of the specific collocated picture;
derive one or more temporal motion vector predictors (TMVPs) based on the plurality of reference motion ve

Assignees

• Mediatek Singapore Pte Ltd
• Zhang Kai
• An Jicheng

Inventors

• Zhang Kai
• An Jicheng

Classifications

• H04N19/30

  using hierarchical techniques, e.g. scalability (H04N19/63 takes precedence) · CPC title

• H04N19/503

  involving temporal prediction (adaptive coding with adaptive selection between spatial and temporal predictive coding H04N19/107; adaptive coding with adaptive selection among a plurality of temporal predictive coding modes H04N19/109) · CPC title

• H04N19/70

  characterised by syntax aspects related to video coding, e.g. related to compression standards · CPC title

• H04N19/52 Primary

  by predictive encoding · CPC title

• H04N19/31

  in the temporal domain · CPC title