An Improved Decoding Algorithm For DVC Over Wireless Transmission

When it occurs to video applications,such as wireless video communication,video surveillance network and wireless video sensor network.due to limited terminal power,video coding scheme is usually required the lowest complexity scheme.However,because traditional video coding techniques(such as MPEG-X and H.264)adopt the techniques of motion prediction and compensation with great complexity,it is difficult to meet the above video application coding' requirements.As a new Video Coding framework,Distributed Video Coding(DVC)moves the motion estimation and compensation operation from the code side to the decode side to meet the requirements of low-complexity Coding under the condition of guaranteeing the compression efficiency.Therefore,DVC is suitable for the above wireless video application and can be regarded as the wireless video transmission system based on DVC.Due to DVC encode scheme uses channel coding technology to replace source coding,so the DVC scheme has the nature ability of error resilient.In order to explore the noise resilient ability of DVC decoding in the wireless transmission environment,this dissertation studies the DVC optimization decoding algorithm based on wireless transmission in the platform of the wireless video transmission system based on DVC,aiming to improve the noise resilient performance of DVC decoding in the wireless environment.The specific research results are as follows:First,with the foundation of Distributed Video Coding system,this dissertation expects to build the wireless video transmission simulation system based on DVC.And try to improve the decode algorithm on this system in order to improve the quality of video transmission.Main work are as follows:Firstly,a simulation system of DVC based wireless video transmission is built on the foundation of distributed video coding.This system is mainly composed of DVC encoder,Convolutional code encode,WCDMA wireless channel,Rake receiver module,Convolutional code decoder and DVC decoder.By using Rake diversity receiver to reduce the Fading Effect in wireless transmission and by using Convolutional codec as error correction means to reduce the distortion effect in the process of video transmission.The experimental results showed that the encoding rate of LDPC in DVC system was reduced from 1/2 to 1/3,and the Peak Signal to Noise Ratio(PSNR)of video sequence foreman,hall monitor and football was increased by 2.32db on average.These results show that video reliability can be improved by reducing the encoding rate of DVC system.Then,with the foundation of wireless video transmission simulation system based on DVC,an improved decoding algorithm for distributed video coding for enhanced performance over error prone wireless channels with multipath fading.This algorithm using the mutual independent of noise distribution of the parity check information with the system information as a break point,and then building models of the parity bit sequence and the system sequence received through a multipath fading wireless channel.And then using the model to influence the posteriori probability of sending side's information,in order to influence the Conditional likelihood ratio in the process of decoding.Experimental results show that when the rates of DVC source code is fixed,the quality of reconstructed video obtained by using the DVC optimized decoding algorithm based on wireless transmission proposed in this dissertation is better than that without optimized decoding algorithm,and the average PSNR increases by 6.4dB.In the case that SNR does not change,with the increase of fading path,the PSNR gain of video decoding reconstruction is maintained above 0.5db by using DVC optimal decoding algorithm in this dissertation.