The Research Of Distributed Video Coding Based On Compressive Sensing

Distributed Video Coding (DVC) is a new coding paradigm based on two majorInformation Theory results: the Slepian-Wolf (1973) and Wyner-Ziv Theorems (1676). Thisnew video coding paradigm allows exploiting the source statistic, partially or totally, at thedecoder only. A particular case of distributed video coding is the Wyner-Ziv video coding. Inthis scenario, two correlated sources are independently encoded using separated encoders andthe bit streams associated to each source are jointly decoded exploiting the correlationbetween them.Although the distributed coding study dates back to the1970’s, efforts towardsdeveloping practical solutions of Wyner-Ziv video coding are more recent. Emergingapplications (such as wireless, low-power surveillance systems and mobile camera phonesamong others) with encoding requirements quite different from those targeted by the MPEG-xand H.26x video coding standards have stimulated such efforts. In the MPEG-x and H.26xstandards, the correlation between temporally adjacent frames is exploited through a complexmotion estimation task which leads to a high complexity encoder. Since the correlationbetween temporally adjacent frames in Wyner-Ziv video coding is performed only at thedecoder, the encoder can typically present a low complexity. Improved error resilience isanother major functionality of this new video coding paradigm since the usual encoderprediction loop and the associated error propagation do not exist anymore. Distributed videocoding are different with the traditional video coding technology. The traditional technologystandards are most often used for fully exploiting the redundant information of video signal inthe encoder, so the coding complexity is generally5~10times as much as decodingcomplexity, it fit for coding one time and decoding many times to video signal (such as videobroadcast, video on demand, video optical disc storage, etc.). Distributed video coding havelow complexity encoder, low power consumption coding and good fault tolerance, it suitablefor the wireless video terminals that their computing power, memory capacity and powerconsumption are limited (such as the wireless video monitoring system, video sensor network,etc.).Firstly, this paper introduces the research status from domestic and abroad of distributedvideo coding, and summarizes some problems. Secondly, introduces the theory basis of distributed video coding, and make a detailed illustration about each part of the codingparadigm. Then it presents a method with theoretical basis which named compressive sensing,and combines the method and distributed video coding to a new paradigm that the encoderand the decoder are improved. At last, uses MATLAB software to build up the two kind ofparadigm, and demonstrates their performances through simulation. The results show that thepractices and the theory are equal. Finally we look to the future about the further practicalityand development trend of the distributed video coding based on compressive sensingtechnology.