Distributed Video Coding Based on Conditional Entropy Hash

Presently, distributed video coding (DVC) arouse high research interests due to the distinct contribution to the low-complexity encoding. In this paper, we propose a hash-based DVC, where the hash bits are compressed by a rate-adaptive low density parity coding accumulate (LDPCA) and sent to aid the decoder in motion estimation. The hash bit-rate is ideally near to its conditional entropy, so we name it conditional entropy hash (CEH). Experimental results demonstrate the CEH-based DVC achieves more improvement in PSNR than the non-conditional entropy hash (NCEH) DVC.