Research On Side Information Estimation Algorithms In Distributed Video Coding Based On Block-based Motion Compensated Temporal Interpolation

With the development of mobile communications, wireless networks and video technology, many new applications arise, such as wireless PC cameras, mobile video phones and video sensor networks. These applications require low power and low complexity encoding. However, the popular digital video coding standards, such as ISO/IEC MPEG and ITU-T H.26x, which exploit complex motion estimation at the encoder, are not well suited for these new applications. As a new video coding paradigm, distributed video coding(DVC) is proposed accordingly. DVC system is an asymmetric video compression scheme where individual frames are encoded independently (intraframe encoding), but decoded conditionally(interframe decoding), to realize the low complexity encoding.In DVC system, side information is of great importance and is one of the standing out difficulties in DVC system. The accuracy of side information would essentially influence the rate distortion and compression performance of a DVC system. The more accurate the side information is, the fewer parity bits is needed to be sent to the decoder, in this way, DVC system has a better compression performance.This paper firstly reviews the theoretic basis and the research development in DVC, and introduces three classical DVC archectures. Then, the paper analyses several general algorithms of side information generation and of multiple side information generation. And the paper then focuses on analyzing and implementing two methods with spatio-temporal side information. The paper mainly proposes a motion compensation based side information estimation algorithm considering block boundary continuity. The proposed algorithm obtains five interpolated frames by motion compensation using bidirectional motion vectors through motion vector smoothing techniques according to different ratios. In this way, motion vectors are more correct, and attain higher quality of interpolated frames. The proposed algorithm also takes advantage of the complementary frames and block boundary continuity to further improve the quality of final interplated frame. Extensive simulations over various test video sequences show, compared with other existing algorithms, the proposed scheme improves the coding efficiency and subjective visual quality of the decoded image.