Wyner-Ziv coding based on signal denoising technique

In Wyner-Ziv video coding, efficient compression is achieved by exploiting source statistics at the decoder only, which is radically different from conventional video coding. Prior work on this topic has been restricted to a basic framework based on algebraic channel coding principles. In this paper, we propose a novel distributed source coding method based on the signal denoising and channel coding approaches that exploit denoising technique to separate the correlation portion of data which is encoded by channel codec, and then encodes the other portion using entropy codec. Our contributions are mainly as follows: (1) To utilize signal denoising and channel coding techniques to implement distributed source coding, a simple correlation structure between the source and the side information is given. (2) A new distributed source coding framework is proposed. (3) To apply the proposed framework in video coding systems, we propose a novel distributed video coding method. Our experimental results show that, comparing to prior distributed video scheme, the proposed scheme can achieve comparable or even better compression efficiency while coding at the same bit rate.