A rate-distortion adaptive order of bitplanes decoding for distributed video coding

Distributed Video Coding (DVC) is a one of the widely researched video compression paradigms for encoder complexity reduction. One of the most popular DVC schemes was proposed in Stanford university. The main disadvantage of Stanford DVC scheme is a feedback channel used by decoder to request parity or syndrome bits. The order of bits requesting can impact on decoding performance significantly. The main idea of the proposed rate-distortion-aware method is to provide fast Wyner-Ziv (WZ) frame reconstruction using minimal number of parity or syndrome bits requested from encoder side. This result is achieved via band and bitplanes re-ordering in request procedure. The proposed approach uses rate-distortion estimation at decoder side to determine the next band/biplane which allows to achieve high(est) error decreasing between original WZ frame and current version of recovered frame at decoder side. The comparison results demonstrate the effectiveness of the developed algorithm.