Adaptive disparity and motion estimation for Multiview Video Coding

Abstract

In the joint mode of MVC (Multiview Video Coding), to fully exploit both temporal and inter-view correlation, motion estimation (ME) and disparity estimation (DE) have been employed to achieve the highest coding efficiency. However, it causes enormous computational complexity. An adaptive ME and DE algorithm is proposed to reduce the complexity in this paper. Fist, we calculate the difference between the coding macroblock (MB) and the corresponding MB in the reference frame to determine the region homogeneity. Then, based on the region homogeneity, some INTER modes will be skipped in coding mode selection. Finally, an optimization method of obtaining the optimal motion information from the neighboring view is proposed to speed up rate distortion optimization (RDO) process for Motion Skip (MS) mode. In addition, the search range (SR) for ME and DE can be limited to an estimable range and the disparity search is selectively enabled adaptively. Experimental results demonstrate that the proposed algorithm can save about 78% computational complexity on average, with negligible loss of encoding efficiency.