The Implementation Of Transcoding From Distributed Video Coding To H.264

DVC (Distributed Video Coding) algorithm based on Slepian-Wolf (1973) lossless distributed coding theory and the Wyner-Ziv (1976) theory of lossy distributed coding, consistent with information theory, has a solid foundation in the mobile terminals and other equipment with relative weak ability of hardware computing. H.264 is formally promulgated at the international video coding standard by the ITU-T in March 2003. With outstanding performance in compression, H.264 is widely used in digital TV broadcasting, video, real-time communication, network video streaming and other areas.The basic characteristics of DVC is that it discards the complicated inter prediction, and thus its calculation is relatively simple, not time-consuming and needs not too many demands on the hardware environment, but its coding compression rate is relatively general; On the other hand, the encoding process of H.264 has a large number of applications in interframe predictive coding algorithm, so it is time-consuming and needs many demands on the hardware environment. Fortunately, the coding compression rate of H.264 is more excellent. Because of these reasons, the transcoding from the DVC to H.264 can both get good performance in time and space, which in reality has very important significance.In this paper, there is a research on the implementation of the transcoding from DVC to H.264. Firstly, the paper introduces the development of video coding, the history of DVC and H.264, the current research on the transcoding from DVC to H.264 at home and abroad, the entire system architecture of the transcoding project, and so on. Secondly, the H.264 video coding algorithm will be studyed at the basic principles of encoding and decoding, intra prediction, inter prediction and the H.264 syntax and semantics, and the paper especially focuses on the inter prediction involving to the mode of the prediction and motion vector. Then the key technologies of DVC, for example, the basis of information theory which DVC depends on, the implementation of trancoding in transform domain and pixel domain, the side information decoding, is analyzed in detail. Followly, each module of the trancoding project is described in detail, especially their implementation. Then the paper focuses on the experiments of the transcoding, assesses the transcoding performance in four different hardware and software environment. Fortunately, the experimental results show that the implementaion of transcoding from DVC to H.264 which the paper provided has a good performance. Finally, the paper summarizes its research work and prospects for the future work at the trancoding from DVC to H.264.