一种基于H.264的帧间错误掩盖算法  

Inter-frame Method of Video Error Concealment in H.264

作　　者：刘冀伟[1] 梁波[1] 王志良[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《小型微型计算机系统》2007年第10期1851-1854,共4页Journal of Chinese Computer Systems

基　　金：教育部科学技术重点研究项目(03021)资助.

摘　　要：H.264采用混合编码算法.视频流的传输错误将会导致解码错误在时间和空间上扩散,直接影响重构图像的质量.本文充分结合H.264的编码特性,利用宏块的空间或时间相关性来确定丢失宏块的类型,根据相邻块的运动矢量建立一个描述小区域内运动矢量变化趋势的多项式模型,更精确的恢复出丢失宏块的运动矢量,提高重构图像的质量.仿真结果表明:该算法优于以往的边界匹配算法,在同等宏块丢失率下,恢复的视频序列Y分量的PSNR值能提高0.75～1.5dB.H. 264 is based on mix encoding frame. Transmission error of video frequency will result in decoding mistake both in time and space, and directly affect the quality of reconstructed image. A new error concealment method is introduced by this paper with sufficient consideration of characteristics for H. 264. With the strong correlation between neighborhood macroblocks, the type and size selection of the lost block is estimated rightly through the neighborhood blocks which have been received or recovered exactly. On the basis of above division result, an efficient motion vector recovery method is proposed. It is based on a polynomial model, which can describe the change tendency of motion vectors within a small area. Then motion vectors of the lost maerobloek could be predicted more exactly, and the quality of the reconstructed image is improved. The simulation results show that this algorithm is better than usual boundary matching arithmetic. Under equal moeroblock lost rate, the Y weight PSNR of video recovered by this proposed method could increase 0. 75-1.5dB.

关 键 词：错误掩盖 边界匹配 多项式模型 PSNR 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]