Multiview video coding based on rectified epipolar lines
Jie Yang, Guang Jiang, Lei Zhang, Geng Zhang, Chengke Wu · 2009
Fast and efficient disparity estimation and compensation are the key techniques in multi-view video processing. When implementing multi-view video coding, there are lots of temporal and inter-viewpoint redundancy among viewpoints, which makes the compression work more challenging. It is found that the inter-viewpoint relationship is subject to the epipolar geometry, which provides constraints between the two image sequences of multi-view video. To efficiently encode data-intensive multi-view imaging content, a fast disparity estimation (DE) search method is proposed in this paper based on the rectified epipolar lines. Theoretical analysis and experimental results prove that the proposed disparity estimation search method can efficiently confine the search range in the process of disparity estimation. Our algorithm can reconstruct images with similar quality compared with the full search and the TZ fast search methods in JMVC (Joint Multi-view Video Coding), but the total encoding time is reduced by 30% and 40% respectively.