QQ客服 官方微博 反馈留言

	Error Concealment with Multiscale Patch Clustering and Low-Rank Minimization
	Li, Mading ; Yang, Wenhan ; Song, Sijie ; Guo, Zongming
	2015
关键词	MATRIX COMPLETION
英文摘要	In this paper, we propose a novel error concealment method based on multiscale patch clustering and low-rank minimization. In order to collect more reliable patches to form a genuine low-rank matrix, an image pyramid is formed utilizing an effective down-sampling process. The classic singular value thresholding (SVT) is modified into a global iteration to solve the low-rank minimization problem. Extensive experimental results on the random pixel loss and the block loss situation validate the effectiveness of the proposed method. The proposed method acquires higher PSNR and better visual quality than the state-of-the-art low-rank based error concealment methods.; CPCI-S(ISTP); 672-675
语种	英语
出处	Asia-Pacific-Signal-and-Information-Processing-Association Annual Summit and Conference (APSIPA ASC)
内容类型	其他
源URL	[http://ir.pku.edu.cn/handle/20.500.11897/450144]
专题	信息科学技术学院
推荐引用方式 GB/T 7714	Li, Mading,Yang, Wenhan,Song, Sijie,et al. Error Concealment with Multiscale Patch Clustering and Low-Rank Minimization. 2015-01-01.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。