Research On Rate Control Technologies For Unidirectional Distributed Video Coding

With the application of video sensor networks, video surveiliance and etc, a novel video coding paradigm, Distributed Video Coding(DVC), is paid great attentions. DVC encodes the video frames in a distributed way, while jointly decodes them by exploiting temporal correlation at the decoder. A great bulk of computation is shifted from encoder to decoder, thus comparable compression performance could achieved with much lower encoding complexity. However, most of DVC architectures in the literature use feedback channels to control rate at the decoder, which would cause great latency and extremely high decoder complexity, thus become a great hindrance to practical application fields without feedback channels.To solve these problems, this paper aims to mainly research rate control technologies in Unidirectional DVC(UDVC) without feedback channels. Firstly, the paper proposes a side information estimation method accurately at encoder, which exploits low-complexity motion estimation algrithom. Then the decoding rate is estimated based on correlation between sideinformation and original information. Secondly, an iterative decoding method without feedback based on side information refinement is introduced at decoder in order to improve RD performance, which decodes probable error bitplans in an iterative decoding way based on reconstruction frames and refined side information. Finally, an UDVC encoding/decoding system is presented based on the above technologies, and experiments were done by practical videos to test the performance of the proposed system. The experiment results show that compared with the bidirectional DVC, the proposd UDVC system could obtain similar quality of the decoded video in a much lower complexity way.