分布式视频编码中关键帧编码方式的改进  

Improvement of key frames in distributed video coding

作　　者：刘晓文[1] 袁莎莎[1] 仲亚丽[1] 

机构地区：[1]中国矿业大学物联网(感知矿山)研究中心,江苏徐州221008

出　　处：《计算机应用研究》2014年第2期619-621,624,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(60972059)

摘　　要：在分布式视频编码中关键帧间的时间相关性没有被利用，据此提出一种基于频带划分的关键帧编码模式的选择方法。关键帧经过DCT后得到的系数带划分为低频带和高频带，低频带部分采用Wyner—Ziv编解码，然后对高频带进一步进行编码模式的选择，分别进行Wyner．Ziv编解码和传统的帧内编解码，且在解码端完成编码模式的选择，通过反馈信道向编码端传送编码模式的选择。仿真结果表明，在不增加编码端复杂度的情况下，所提出的方法改善了解码后视频的质量，使系统的峰值信噪比（PSNR）提高约1。5dB。In most distributed video coding approaches, the temporal correlation of the key frames is not exploited. This paper proposed a key frames coding mode selection method based on band allocation. Firstly, it divided coefficient tape obtained from DCT of the key word into low frequency and high frequency, applied Wyner-Ziv codec to the low frequency. Secondly, it selected an appropriate coding mode for the high frequency, taking Wyner-Ziv codec and traditional intra-frame codee respec- tively. It completed the selection of the key frames coding mode in decoder, then, the decoder transformed the selection results to the encoder through the feedback channel. Simulation results show the peak signal to noise ratio （PSNR） of the systems in- creases by 1 to 5rib and the quality of decoded video is improved without increasing the complexity of the encoder.

关 键 词：无线多媒体传感器网络 变换域Wyner—Ziv 关键帧 离散余弦变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]