Research On Video Compression Coding And Reliable Transmission For WMSNs

Wireless multimedia sensor networks has predictive application in military field, commerce domain, and security. Motion estimation encoding algorithm, its low-complexity realization, and reliable transmission are the key points when WMSNs come to utility. This thesis analyzed video compression algorithm and reliable transmission suiting WMSNs. The main work and innovation are as following:(1) H.264 video compression standard was analyzed after the comparison of existing video compression standards. Its encoding performance, especially the characteristics and advantages were discussed. The key issues, such as inner-coding, intra-coding, motion compensation and estimation based on block were researched. Meanwhile, the principle of classical fast search algorithms: Trivial Step Search, Four Step Search, Diamond Search, and newly proposed fast search algorithms: Hexagon Based Search, DHCDDS, UMHexagonS were discussed here.(2) A new motion estimation coding method was researched. Two aspects were considered to reduce computation complexity. The relativity of current block and its left, above, above right neighbors was used to estimate the search start point. When the relative position is zero vector, direct replacement or straight search was adopted according to the difference scope. As to nonzero prediction vector, the best search beginning was derived from weighed average. Compressed sensing was used to find the best matching macroblock. A resolution algorithm for motion vector was gained using measure matrix basis, which was reconstructed from current macroblock coefficients. When the precision reached the request threshold, search came to the end ahead of schedule.(3) considered the limited energy and computation ability of WMSNs, low-complexity realization of video compression technology and code optimization were researched. Integer transform and fixed-point LBT were used to avoid error caused by floating-point operation and actualize algorithm in low complexity. Taking the difference between PC and ARM into account, optimization method for replanting code was designed.(4) Aiming at the instability of wireless channel, error detection and source semantic error concealment algorithm was researched. A block check error orientation method was advanced after analysis of syntax and semantic based error detection. And then a source semantic error concealment algorithm was proposed using candidate slice selected in compressed sensing faint matching method.