Advanced merge mode for three-dimensional (3D) video coding

Patent metadata

Field	Value
Publication number	US-9609347-B2
Application number	US-201414243610-A
Country	US
Kind code	B2
Filing date	Apr 2, 2014
Priority date	Apr 4, 2013
Publication date	Mar 28, 2017
Grant date	Mar 28, 2017

Abstract

As part of a video encoding process or a video decoding process, a video coder may determine a first available disparity motion vector among spatial neighboring blocks of a current block of the video data. Furthermore, the video coder may shift a horizontal component of the first available disparity motion vector to derive a shifted disparity motion vector candidate (DSMV). The video coder may add the DSMV into a merge candidate list.

First claim

What is claimed is:

1. A method of decoding video data, the method comprising:
determining, by a video decoder, a shifted disparity vector by shifting a disparity vector of a current block of the video data horizontally by a first value and vertically by a second value, the first value being equal to two times the width of the current block, and the second value being equal to two times the height of the current block;
using, by the video decoder, the shifted disparity vector to generate an inter-view prediction motion vector candidate (IPMVC) indicating motion information of a corresponding block in an inter-view reference picture in a reference view, the corresponding block corresponding to the current block;
responsive to determining the IPMVC is unavailable, adding, by the video decoder, a disparity shifted motion vector (DSMV) candidate into a merge candidate list, wherein a horizontal component of the disparity vector is shifted by a third value to derive a motion vector of the DSMV candidate, the third value being equal to 4;
determining, by the video decoder, based on an index signaled in a bitstream, a selected merge candidate in the merge candidate list;
generating, by the video decoder, a predictive block for the current block using motion information of the selected merge candidate; and
adding, by the video decoder, residual data to the predictive block to reconstruct a block of the video data.

2. The method of claim 1 , wherein:
the motion vector of the DSMV candidate is a first motion vector of the DSMV candidate, and the method comprises, based on at least one of a plurality of spatial neighboring blocks of the current block having a disparity motion vector:
setting, by the video decoder, the first motion vector of the DSMV candidate to a first motion vector of a particular spatial neighboring block among the plurality of spatial neighboring blocks of the current block, the first motion vector of the particular spatial neighboring block corresponding to a first reference picture list, and the first motion vector of the particular spatial neighboring block being a first available disparity motion vector among the plurality of spatial neighboring blocks;
setting, by the video decoder, a second motion vector of the DSMV candidate to a second motion vector of the particular spatial neighboring block, the second motion vector of the particular spatial neighboring block corresponding to a second reference picture list;
setting, by the video decoder, a reference index corresponding to the first motion vector of the DSMV candidate to a reference index corresponding to the first motion vector of the particular spatial neighboring block;
setting, by the video decoder, a reference index corresponding to the second motion vector of the DSMV candidate to a reference index corresponding to the second motion vector of the particular spatial neighboring block; and
adding, by the video decoder, the third value to the horizontal component of the first motion vector of the DSMV candidate.

3. The method of claim 1 , wherein the motion vector of the DSMV candidate is a first motion vector of the DSMV candidate, the method further comprising, based on none of a plurality of spatial neighboring blocks of the current block having a disparity motion vector:
setting, by the video decoder, the first motion vector of the DSMV candidate and a second motion vector of the DSMV candidate to the disparity vector of the current block;
adding, by the video decoder, the third value to the horizontal component of the first motion vector of the DSMV candidate and the third value to a horizontal component of the second motion vector of the DSMV candidate; and
setting, by the video decoder, a vertical component of the first motion vector of the DSMV candidate and a vertical component of the second motion vector of the DSMV candidate to 0.

4. The method of claim 3 , further comprising:
performing, by the video decoder, a process based on neighboring blocks to determine the disparity vector of the current block, wherein the process identifies the reference view along with the disparity vector of the current block; and
based on none of the spatial neighboring blocks having a disparity motion vector:
setting, by the video decoder, a reference index corresponding to the first motion vector of the DSMV candidate to a reference index of a picture that is in a first reference picture list and that belongs to the identified reference view; and
setting, by the video decoder, a reference index corresponding to the second motion vector of the DSMV candidate to a reference index of a picture that is in a second reference picture list and that belongs to the identified reference view.

5. The method of claim 2 , wherein based on backward warping view synthesis prediction being enabled for a whole picture or a current sequence, performing at least one of:
setting, by the video decoder, a vertical component of the first motion vector of the DSMV candidate to 0 based on the first motion vector of the DSMV candidate being derived from the first available disparity motion vector; and
setting, by the video decoder, a vertical component of the second motion vector of the DSMV candidate to 0 based on the second motion vector of the DSMV candidate being derived from the first available disparity motion vector.

6. The method of claim 1 , wherein adding the DSMV candidate into the merge candidate list comprises: adding, by the video decoder, the DSMV candidate into the merge candidate list immediately after a merge candidate for a spatial neighboring block that is above and left of the current block.

7. The method of claim 1 , wherein the current block is a first block and the corresponding block is a first corresponding block, and the IPMVC is a first IPMVC, the method further comprising:
using, by the video decoder, a disparity vector of a second block to locate a second corresponding block in the reference view, the second corresponding block corresponding to a second block;
responsive to determining that motion information of the second corresponding block is available, inserting, by the video decoder, a second IPMVC into a merge candidate list for the second block, the second IPMVC indicating at least one motion vector of the second corresponding block;
determining, by the video decoder, a second shifted disparity vector by shifting the disparity vector of the second block horizontally by a fourth value and vertically by a fifth value;
using, by the video decoder, the second shifted disparity vector to locate a third corresponding block in the reference view, the third corresponding block corresponding to the second block, and
responsive to determining that motion information of the third corresponding block is available and does not match motion information of the second IPMVC, inserting, by the video decoder, a third IPMVC into the merge candidate list for the second block, the third IPMVC indicating at least one motion vector of the third corresponding block.

8. The method of claim 7 , wherein inserting the third IPMVC into the merge candidate list for the second block comprises:
inserting, by the video decoder, the third IPMVC into the merge candidate list for the second block immediately after a merge candidate for a spatial neighboring block that is above and left of the second block.

9. An apparatus for decoding video data, the apparatus comprising:
a memory configured to store the video data, and
one or more processors configured to:
determine a shifted disparity vector by shifting a disparity vector of a current block of the video data horizontally by a first value and vertically by a second value, the first value being equal to two times th

Assignees

• Qualcomm Inc

Inventors

• Thirumalai Vijayaraghavan
• Zhang Li
• Chen Ying

Classifications

• H04N19/513 Primary

  Processing of motion vectors · CPC title

• H04N19/56

  Motion estimation with initialisation of the vector search, e.g. estimating a good candidate to initiate a search · CPC title

• H04N19/70

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

• H04N19/597

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