首页 | 本学科首页   官方微博 | 高级检索  
     

基于MRF-MAP的运动矢量恢复差错掩盖算法
引用本文:许志良,王建华,谢胜利.基于MRF-MAP的运动矢量恢复差错掩盖算法[J].计算机工程,2008,34(14):222-224.
作者姓名:许志良  王建华  谢胜利
作者单位:1. 深圳信息职业技术学院,深圳,518029
2. 华南理工大学电子与信息学院,广州,510641
基金项目:国家自然科学基金 , 广东省深圳市科技计划 , 深圳信息职业技术学院校博士创新基金
摘    要:针对帧间编码模式下出现图像块丢失的情况,提出一种有效的时间域运动矢量恢复差错掩盖算法。把运动矢量场建模为高斯马尔科夫随机场,对丢失图像块的运动矢量采用最大后验概率方法恢复,其权值能够根据空间和时间信息而自适应选择。仿真结果表明,该算法在客观和主观上都能获得高质量的图像。

关 键 词:运动矢量  马尔可夫随机场  最大后验概率  差错掩盖
修稿时间: 

Error Concealment Algorithm for Motion Vector Recovery Based on MRF-MAP
XU Zhi-liang,WANG Jian-hua,XIE Sheng-li.Error Concealment Algorithm for Motion Vector Recovery Based on MRF-MAP[J].Computer Engineering,2008,34(14):222-224.
Authors:XU Zhi-liang  WANG Jian-hua  XIE Sheng-li
Affiliation:(1. Shenzhen Institute of Information Technology, Shenzhen 518029; 2. College of Electronic & Information Engineering, South China University of Technology, Guangzhou 510641)
Abstract:Error concealment is an attractive approach to combat channel errors for video transmission. Aiming at recovering the damaged macroblocks which belong to inter-coded model, an effective error concealment algorithm for motion vector recovery based on MRF-MAP is proposed in this paper. The motion vectors field is modeled as Gauss-Markov Random Field(GMRF). The motion vectors of the damaged image macroblocks can be recovered adaptively by Maximum A Posteriori(MAP), and the weight is selected adaptively based on the spatial information and temporal information. Simulation results show that the proposed method offers significant improvement on both objective PSNR measurement and subjective visual quality of restored video sequence.
Keywords:motion vector  Markov Random Field(MRF)  Maximum A Posteriori(MAP)  error concealment
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机工程》浏览原始摘要信息
点击此处可从《计算机工程》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号