Wyner-Ziv视频系统中解码算法研究  

Study of Decoding Algorithm for Wyner-Ziv Video Coding System

作　　者：干宗良[1] 齐丽娜[1] 朱秀昌[1] 

机构地区：[1]南京邮电大学图像处理与图像通信重点实验室,南京210003

出　　处：《信号处理》2008年第4期609-613,共5页Journal of Signal Processing

基　　金：江苏省自然科学基金(BK2004151)

摘　　要：首先简要介绍了一种典型的分布式视频编码-Wyner-Ziv视频编码。然后对Wyner-Ziv视频编码中边信息进行理论分析,随后给出了基于加权MAD准则的边信息估计算法和基于先验概率约束的联合解码算法。实验仿真结果表明,采用本文解码优化策略,在编码端相同输出码率情况下,重建解码图像的PSNR比原始算法平均提高1.5dB。The Wyner-Ziv video coding,which is a kind of typical distributed video coding scheme,is introduced at first. Then the mathematical analysis of the side information in Wyner-Ziv video coding is given and we propose an improved decoding algorithm of Wyner-Ziv video coding based on side information estimation using weighted MAD motion estimation criterion and decoding reconstruction based on the prior model. The results of simulation tests show that average 1.5dB gain can be achieved by applying the proposed method.

关 键 词：分布式视频编码 Wyner—Ziv视频编码 边信息 运动估计 

分 类 号：TN911.22[电子电信—通信与信息系统] TN948.12[电子电信—信息与通信工程]