Distributed source coding: Theorem and its application to video coding

Distributed Source Coding (DSC) implements inter source redundancy by encoding correlated information sources, without letting the sources communicate with each other. The DSC theorem states that the optimal rate achieved by joint encoding and decoding can be reached by separate encoding and joint decoding. Video coding based on the DCS theorem is Distributed Video Coding (DVC). DVC is attracting attention as a new paradigm for video compression. While conventional video coding schemes such as H.264/AVC reduce inter-frame redundancy at the encoder, DVC reduces inter-frame redundancy at the decoder by using distributed channel coding. The architecture suits applications that require low-complexity encoders. Examples of such applications include wireless video surveillance and mobile camera phones. In this paper, after describing the DSC theorem, we offer a survey of recent trends and problems in DVC and then introduce our contributions.