H.264 Time-domain Variable Size Block-based Motion Vector Recovery For Error Concealment Algorithm

With the information age developing in the word, the application of multimedia has penetrated every area in human life. The video have won the care and favor of people due to its lively, visual and abundant content. On the other hand, its huge information limits its flexible application. Human deals with this problem by video compress technology partly Video conference, Video disc and so on come into people's life gradually. But with the new application's emergence of IPTV and mobile Video, these previous video compress technologies become more and more difficult to suffice so many demands of above business over video application gradually.Therefore, ITU-T and ISO/IEC , With the tenet to design a simple and effective new codec technology, with the goal to get outstanding compress performance and excellent network adaption, issued the H.264 video compress standard in May 2003. Being the latest standard in video compress field, the standard makes it possible to implement many new and future video applications. Through inheriting, ameliorating and innovating, H.264 synthesizes so many advantages of the previous standard, which not only enhances its compress performance, but also gets eminent predominance noise in transmission over channel, in contrast with previous video compress standard; the influence of noise will make more serious effect over H.264 bit stream. Thereby after detailed introducing H.264 flexible application. Human deals with the latest video coding standard H.264 supports variable size block motion prediction and compensation. Adapt to this new feature, a new temporal error concealment algorithm based on variable size block motion vector resilience, is proposed in this paper. Exploiting the correlation of adjacent macro blocks, the encoding mode of lost macro block, partition pattern of lost macro block, and the reference frame used in matching can be decided according to neighboring macro blocks' encoding mode and used reference frame information, and then motion vector is recovered for each partitioned sub blocks using side distortion matching method. Simulation results show the proposed algorithm is superior to conventional temporal concealment methods for H.264 standard.