作者: Jian-Guang Lou , Hua Cai , Jiang Li
DOI:
关键词:
摘要: Described herein is technology for, among other things, multiview coding with geometry-based disparity prediction. The prediction involves determining corresponding block pairs in a number of reconstructed images for an image being coded. and the represent different views scene at point time. Each pair projected on image. This enables determination vector candidates. For each image, predicted determined based Then, may be utilized to obtain bits encoded. reduces encoded bits.