Advanced merge/skip mode and advanced motion vector prediction (AMVP) mode for 3D video

Patent metadata

Field	Value
Publication number	US-9357214-B2
Application number	US-201314098151-A
Country	US
Kind code	B2
Filing date	Dec 5, 2013
Priority date	Dec 7, 2012
Publication date	May 31, 2016
Grant date	May 31, 2016

Abstract

Techniques are described where if an inter-view predicted motion vector candidate (IPMVC) and an inter-view disparity motion vector candidate (IDMVC) are derived based on a shifted disparity vector, where the amount by which the disparity vector is shifted for the IPMVC and IDMVC is different. The techniques also prioritize the inclusion of the IPMVC over the IDMVC in a candidate list, and prune the IPMVC and the IDMVC if there is a duplicated IPMVC or IDMVC in the candidate list.

First claim

What is claimed is:

1. A method of coding video data, the method comprising:
determining that a current block is inter-predicted in merge mode or inter-predicted in skip mode;
based on the current block being inter-predicted in merge mode or inter-predicted in skip mode:
identifying a reference block in a reference view referred to by a disparity vector, for the current block, shifted by a first offset value;
determining that motion information for the identified reference block is not available;
based on the motion information of the identified reference block being not available, determining that an inter-view predicted motion vector candidate (IPMVC) is not available for the current block and
based on the IPMVC not being available, deriving an inter-view disparity motion vector candidate (IDMVC) and including the IDMVC in a candidate list for the current block, wherein the IDMVC comprises the disparity vector, for the current block, shifted by a second offset value that is different than the first offset value used to shift the disparity vector for identifying the reference block; and

inter-prediction coding the current block in a current picture in a current view, different form the reference view, based on the candidate list.

2. The method of claim 1 , wherein the current block comprises a first block, the disparity vector comprises a first disparity vector, the candidate list comprises a first candidate list, the reference block comprises a first reference block, and the reference view comprises a first reference view, the method further comprising:
identifying a second reference block in a second reference view referred to by a second disparity vector, for a second block that is inter-predicted in merge mode or inter-predicted in skip mode, shifted by a third offset value;
determining that an IPMVC for the second block is available based on a motion vector for the second reference block in the second reference view referred to by the second disparity vector shifted by the third offset value being a temporal motion vector that refers to a picture in the second reference view with a picture order count (POC) value that is equal to a POC value of a picture in a reference picture list for the second block
deriving the IPMVC for the second block based on the IPMVC being available;
avoiding the deriving of an IDMVC for the second block based on the IPMVC not being available, wherein IDMVC for the second block comprises the second disparity vector, for the second block, shifted by a fourth offset value that is different than the third offset value used to shift the second disparity vector for identifying the second reference block;
including the IPMVC for the second block in a second candidate list for the second block; and
inter-prediction coding the second block based on the second candidate list for the second block.

3. The method of claim 1 , wherein the current block comprises a first block, the disparity vector comprises a first disparity vector, the candidate list comprises a first candidate list, the reference block comprises a first reference block, and the reference view comprises a first reference view, the method further comprising:
determining that a reference picture index for a second block, that is inter-predicted in advance motion vector prediction (AMVP) mode, identifies one of a temporal reference picture or an inter-view reference picture;
based on the reference picture index identifying the temporal reference picture:
deriving an IPMVC for the second block, wherein the IPMVC for the second block comprises motion information for a second reference block in a second reference view referred to by a second disparity vector, for the second block, shifted by a third offset value;
including the IPMVC in a second candidate list for the second block; and
inter-prediction coding the second block based on the second candidate list and based the reference picture index identifying the inter-view reference picture:
deriving an IDMVC for the second block, wherein the IDMVC for the second block comprises the second disparity vector, for the second block, shifted by a fourth offset value that is different than the third offset value that would have been used had the IPVMC been derived;
including the IDMVC in a third candidate list for the second block; and

inter-prediction coding the second block based on the third candidate list.

4. The method of claim 1 , wherein the reference block comprises a first reference block, the current block comprises a first block, the disparity vector comprises a first disparity vector, and the reference view comprises a first reference view, the method further comprising:
determining a first IPMVC for a second block that is inter-predicted in merge mode or inter-predicted in skip mode, wherein the first IPMVC comprises motion information for a second reference block in a second reference view referred to by a second disparity vector, for the second block, shifted by a third offset value;
determining a second IPMVC for the second block, wherein the second IPMVC comprises motion information for a third reference block in the second reference view referred to by the second disparity vector, for the second block, without shifting the second disparity vector; and
comparing the first IPMVC and the second IPMVC;
including the first IPMVC in a candidate list for the second block based on the comparison indicating that the second IPMVC is different than the first IPMVC; and
inter-prediction coding the second block based on the candidate list for the second block.

5. The method of claim 4 , wherein comparing comprises comparing the first IPMVC only with the second IPMVC.

6. The method of claim 1 , further comprising:
determining the disparity vector for the current block based on one of a disparity motion vector for a neighboring block and a derived disparity vector for a neighboring block;
adding the first offset value to or subtracting the first offset value from the disparity vector to determine the disparity vector shifted by the first offset value; and
based on the IPMVC not being available, adding the second offset value to or subtracting the second offset value from the disparity vector to determine the disparity vector shifted by the second offset value.

7. The method of claim 1 , wherein the second offset value comprises a fixed offset value for shifting the disparity vector horizontally.

8. The method of claim 1 , wherein the motion information of the reference block comprises at least one of a motion vector and one or more reference indices for reference picture lists.

9. The method of claim 1 , further comprising:
receiving an index into the candidate list,
wherein determining that the current block is inter-predicted in merge mode, or inter-predicted in skip mode comprises decoding syntax elements that indicate that the current block is inter-predicted in merge mode or inter-predicted in skip mode, and
wherein inter-prediction coding comprises inter-prediction decoding the current block based on information identified by the received index into the candidate list.

10. The method of claim 1 , wherein inter-prediction coding comprises inter-prediction encoding the current block based on information in the candidate list, the method further comprising:
signaling syntax elements that indicate that the current block is inter-predicted in merge mode or inter-predicted in skip mode; and
signaling an index into the candidate list that is used to identify the information in the candidate list.

11. A device for coding video data, the device comprising a video coder configured to:
determine that a current block is inte

Assignees

• Qualcomm Inc

Inventors

• Zhang Li
• Chen Ying
• Karczewicz Marta

Classifications

• H04N19/597

  specially adapted for multi-view video sequence encoding · CPC title

• H04N19/52 Primary

  by predictive encoding · CPC title

• H04N19/103

  Selection of coding mode or of prediction mode · CPC title

• H04N19/00696 Primary

  Electricity · mapped topic