Multi-viewpoint video coding and decoding system, method and device for estimating vector

Abstract

The invention discloses a method for predicting a vector in the multi-view point view coding, comprising the following steps that: the coding way adopted in a current image block is determined; if the current image block adopts the inter-view interframe prediction coding way, the vector of the difference of the view points of the current image block is predicted; if the current image block adopts the intra-view interframe prediction coding way, the motion vector of the current image block is predicted. The invention also discloses a vector predicting device in the multi-view point view coding, which is characterized in that the device comprises a current image block determining unit, a coding way determining unit, a vector of the difference of the view points predicting unit and a motion vector predicting unit. The invention also discloses a method for predicting the vector in the multi-view point view coding and a multi-view point view coding system. The method, the device and the system disclosed by the invention can improve the prediction accuracy of the vector of the difference of the view points and the motion vector and raise the coding efficiency of the multi-view point video.

Description

technical field [0001] The present invention relates to encoding technology, in particular to a multi-viewpoint video encoding and decoding system, and a method and device for predicting vectors in the multi-viewpoint video encoding and decoding system. Background technique [0002] With the development of multimedia communication technology, people are no longer satisfied with traditional fixed viewpoint vision and 2D plane vision. In many application fields such as entertainment, education, tourism and surgery, there is a demand for free viewpoint video and 3D video. Examples include Free Viewpoint Television (FTV), which allows the viewer to select a viewing angle, and Stereoscopic Television (3DTV), which plays videos from different viewing angles for viewers in different positions. The above-mentioned applications all require the use of multiple cameras to obtain video signals of the same scene at different spatial positions and angles at the same time, and effectively ...

AI Technical Summary

Problems solved by technology

[0004] Because, in multi-view video, the vectors of image blocks are divided into motion vectors and viewpoint difference vectors, and there is usually no correlation between motion vectors and viewpoint difference vectors, so, according to the existing technology, it is impossible to accurately predict the The vector of the block, the prediction accuracy of the vector is not high, resulting in a decrease in coding efficiency

Images