Hybrid Side Information Generation Algorithm Based on Probability Fusion for Distributed Video Coding

Abstract

In a distributed video coding (DVC) system, side information(SI) is the estimated frame of the original Wyner-Ziv(WZ) frame, which has an important impact on the performance of the whole system. High quality SI is helpful for the decoder to restore Wyner-Ziv frames. For this reason, based on the analysis of the existing methods of generating SI, we propose a new SI generation method. Two categories of SI are generated by motion compensation frame interpolation(MCFI) and optical flow interpolation, and then they are fused based on a probability fusion method to obtain more accurate SI. The experiment results indicate that the Hybrid SI generation algorithm proposed in this work can effectively improve the quality of SI.