摘要
为了克服视频传输中因传输错误引起的视频质量下降,结合H.264的编码特性,提出一种基于运动矢量自适应修复的时域差错掩盖算法。首先利用正确接收的宏块信息估计丢失块邻域内运动矢量的变化强度,据此自适应采用边界运动矢量差值最小法或多项式插值法恢复丢失块的运动矢量,然后根据恢复的运动矢量对丢失块进行补偿替代。仿真结果表明,该算法同传统的时域差错掩盖算法相比,在相同网络丢包的环境下PSNR值提高了0.2～2.5 dB,同时解码的图像也获得较好的主观视觉效果。
To overcome the degradation of video quality caused by transmission errors, a new temporal error concealment algorithm based on motion vector adaptive recovery was proposed, which efficiently exploited the characteristic of H. 264. Firstly the Motion Vector Relative Strength （MVRS） was estimated by calculating a ratio involving the motion vectors of adjacent macro blocks. Then either minimum variance of boundary motion vectors method or polynomial interpolation method was employed to recover the lost motion vectors according to the magnitude of MVRS. The simulation results show that the proposed algorithm can highly improve the quality of reconstructed video and obtain a gain of about 0.2 - 2.5dB in PSNR, compared with conventional temporal error concealment algorithms in the condition of equal packet loss rate.
出处
《计算机应用》
CSCD
北大核心
2009年第5期1355-1358,共4页
journal of Computer Applications
基金
湖北省信息产业发展专项资金资助项目(2007014)
关键词
视频传输
错误掩盖
时域掩盖
边界匹配
多项式插值
video transmission
error concealment
temporal concealment
boundary matching
polynomial interpolation
