2004 International Conference on Image Processing ( U P) WYNER-ZIV VIDEO CODING WITH HASH-BASED MOTION COMPENSATION AT THE RECEIVER

In current interframe video compression systems, the en-coder performs predictive coding to exploit the similarities of successive frames. The Wyner-Ziv Theorem on source coding with side information availahlc only at the decoder suggests that an asymmetric video codec, where individual frames are errcoded separately, but decoded conditionally (given temporally adjacent frttmes) could achieve similar efficiency. In previous work we propose a Wyner-Ziv cod-ing scheme for motion video that uses intraframe encoding, but iriterfranie decoding. In this paper we improve on our Wyner-Ziv video codec by sending hash codewords of the current frame to aid the decoder in <accurately estiniating the motion. This allows US to iriiplenicnt a low-delay sys-teni where only the previous reconstructed frame is used to generate the side inforniation of a current frame. Simula-tion results show significant gains above conventional DCT-based inlreframe coding. The Wyner-Ziv video codec with h a s h- h a d motion compensation at the receiver enables low-complexity encoding while achieving high compression efficiency. 1