Method of Deriving Default Disparity Vector in 3D and Multiview Video Coding

Patent metadata

Field	Value
Publication number	US-2016182884-A1
Application number	US-201414908273-A
Country	US
Kind code	A1
Filing date	Aug 13, 2014
Priority date	Aug 13, 2013
Publication date	Jun 23, 2016
Grant date	—

Abstract

A method and apparatus for a three-dimensional or multi-view video encoding or decoding system utilizing unified disparity vector derivation is disclosed. When a three-dimensional coding tool using a derived disparity vector (DV) is selected, embodiments according to the present invention will first obtain the derived DV from one or more neighboring blocks. If the derived DV is available, the selected three-dimensional coding tool is applied to the current block using the derived DV. If the derived DV is not available, the selected three-dimensional coding tool is applied to the current block using a default DV, where the default DV is set to point to an inter-view reference picture in a reference picture list of the current block.

First claim

1 . A method of video coding for a three-dimensional or multi-view video encoding or decoding system, the method comprising:
receiving input data associated with a current block in a dependent view; determining a selected three-dimensional coding tool, wherein the three-dimensional coding tool utilizes a derived DV (disparity vector); deriving the derived DV for the current block from one or more neighboring block of the current block; applying the selected three-dimensional coding tool to the current block using the derived DV if the derived DV is available; and if the derived DV is not available, applying the selected three-dimensional coding tool to the current block using a default DV, wherein the default DV is set to point to an inter-view reference picture in one reference picture list of the current block.

2 . The method of claim 1 , wherein the derived DV is available if a first inter-view reference picture in a first view associated with the derived DV is in one reference picture list of the current block.

3 . The method of claim 1 , wherein the selected three-dimensional coding tool corresponds to VSP (view synthesis prediction).

4 . The method of claim 1 , wherein a default view index of the default DV is set to a view index of one inter-view reference picture of a current slice or picture with a minimum view index, and wherein the current slice or picture contains the current block.

5 . The method of claim 1 , wherein a default view index of the default DV is set to a view index of any inter-view reference picture in one reference picture list of the current block.

6 . The method of claim 1 , wherein a default view index of the default DV is set to a view index of one inter-view reference picture of a current slice or picture having a nearest view index, wherein the nearest view index is measured based on view distance or view index difference with the current slice or picture, and the current slice or picture contains the current block.

7 . The method of claim 1 , wherein if the derived DV is not available and if no inter-view reference picture can be found in any reference picture list of the current block, the default DV is set to a default view index corresponding to −1.

8 . The method of claim 1 , wherein if the derived DV is not available and if no inter-view reference picture can be found in any reference picture list of the current block, the selected three-dimensional coding tool is disabled.

9 . The method of claim 1 , wherein a default view index of the default DV is set to a view index of one inter-view reference picture having smallest quantization parameters.

10 . The method of claim 1 , wherein a default view index of the default DV is set to a view index of one inter-view reference picture that is found firstly among a search set according to a search order, wherein the search set includes all inter-view reference pictures in one or two reference lists of the current block, and wherein the search order starts from a zero picture index to a maximum reference picture index.

11 . The method of claim 10 , wherein the search set includes all inter-view reference pictures in reference list- 0 and reference list- 1 of the current block when the current block corresponds to a prediction unit (PU) in a B slice, and wherein the inter-view reference pictures in the reference list- 0 are searched before or after the inter-view reference pictures in the reference list- 1 , or the inter-view reference pictures are searched in an interleaved order between the reference list- 0 and the reference list- 1 .

12 . The method of claim 1 , wherein a default view index of the default DV is set to a view index of one inter-view reference picture that is found firstly among a search set according to a search order, wherein the search set includes all inter-view reference pictures in one or two reference lists of the current block, and wherein the search order starts from a zero view index to current view index minus 1.

13 . The method of claim 12 , wherein the search set includes all inter-view reference pictures in reference list- 0 and reference list- 1 of the current block when the current block corresponds to a prediction unit (PU) in a B slice, and wherein the search order starts from a zero picture index to a maximum reference picture index and wherein the inter-view reference pictures in the reference list- 0 are searched before or after the inter-view reference pictures in the reference list- 1 , or the inter-view reference pictures are searched in an interleaved order between the reference list- 0 and the reference list- 1 .

14 . The method of claim 1 , wherein a vector value of the default DV is set to a zero vector or a default vector.

15 . The method of claim 14 , wherein the default vector is derived from a converted disparity that is converted from a default depth value.

16 . The method of claim 15 , wherein the default depth value is explicitly signaled or implicitly determined for both an encoder and a decoder.

17 . The method of claim 15 , wherein the default depth value is determined based on a middle value, a mean value, or a medium value of valid depth values, or based on a dominant depth value, and wherein the dominant depth value is determined based on statistic of previously reconstructed depth values.

18 . (canceled)

19 . The method of claim 14 , wherein the default vector is set to a selected disparity from default disparity values.

20 . The method of claim 1 , wherein the default DV is determined at each slice level or each picture level.

21 . The method of claim 1 , wherein the derived DV for the current block is derived from one or more spatial neighboring block of the current block, one or more temporal neighboring block of the current block, or one or more spatial and one or more temporal neighboring blocks of the current block.

22 . (canceled)

Assignees

• Mediatek Inc

Inventors

• Lin Jian-Liang
• Zhang Na
• Chen Yi-Wen
• An Jicheng
• Chang Yu-Lin

Classifications

• H04N2013/0081

  Depth or disparity estimation from stereoscopic image signals · CPC title

• H04N19/44

  Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title

• H04N19/593

  involving spatial prediction techniques · CPC title

• H04N13/0048 Primary

  Electricity · mapped topic