Research On Video Coding For Network Transimission

There are two problems in video communication, one is heterogeneity of network, and the other is reliability of transmission. In video coding, layered coding is proposed to solve client heterogeneity problems, and multiple description coding is an effective method for robust transmission. Multiple description layered coding combines advantages of layered coding and multiple description coding.Firstly, we proposed a macroblock splitting multiple description scheme based on block coding and corresponding redundancy and mismatch control methods in this paper. The proposed scheme is compatible with video coding standard, and each description can be decoded by H.263 decoder. Experimental results show that the performance of macroblock splitting is better than that of optimal DCT coefficient splitting and MDTC.Secondly, We analyzed applications of multiple description coding on CDN (Content Delivery Network) in this paper. We also proposed three multiple description coding schemes for CDN, analyzed temporal and spatial relativities of different video sequences, and compared their performance of the three schemes. Experimental results show that the coding efficiency of odd-even frames decomposing multiple description scheme based on temporal relativities is better than that of odd-even columns decomposing and odd-even rows decomposing schemes based on spatial relativities.Finally, a new multiple description layered coder of video is proposed, based on macroblock splitting technique. At the same time, other three multiple description layered coding schemes are also given, based on row-decomposition, column-decomposition, or frame-decomposition, respectively. The experimental results show that the proposed multiple description video layered coding with macroblock splitting has obvious advantages in adaptability of network heterogeneity and transmission reliability.