摘要
当视频传输中出现不能恢复的误码或丢包时,在接收端就要使用误码掩盖技术对损坏的视频进行掩藏。文中提出了一种基于模糊聚类的多特征量软判决误码掩盖方法,使得错误恢复效果提高。提出的算法主要分两个步骤:首先通过选择特征量对丢失块和周围相邻的块进行聚类分析,找出和丢失块相似的块,然后根据相似块的分布情况对丢失块进行掩盖。
When bit errors and packets loss are not recoverable during video transmission, an error concealment scheme is needed to mask the damaged image at receiver. In this paper, we propose a spatial error concealment algorithm based on fuzzy clustering. The algorithm includes two steps: first classifying the neighbor blocks around the damaged MB using fuzzy clustering, and then interpolating the bad block based on the distribution of the similar blocks. The simulation results show that this algorithm improves the quality of error concealment saliently.
出处
《南京邮电大学学报(自然科学版)》
EI
2006年第4期40-43,共4页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
基金
江苏省图像处理与图像通信实验室(KJS2093)资助项目
关键词
模糊集
聚类分析
误码掩盖
线性内插
Fuzzy sets
Classify analysis
Error concealment
Linear interpolation