CASIA OpenIR  > 模式识别国家重点实验室  > 图像与视频分析
An adaptive pointwise P-norm regularization for image deconvolution
Binbin, Ma1,2; Xiyuan, Hu1; Silong, Peng1
Conference NameInternational Conference on Digital Signal Processing
Conference Date2019年4月23日
Conference PlaceJeju Island, Korea

We propose a pointwise P-Norm based nonparametric image deconvolution method. In our algorithm, a pointwise p-norm based regular term is used to deal with the different types of regions in a blurry image such as edges, texture and smooth areas. A new hyper-parameter updating method is also adopted to improve the effectiveness and robustness of the proposed algorithm. The experimental results show that the quality of our restored images can outperform the results derived by other algorithms. In addition, our algorithm has a wide adaptation to different blur kernels.

KeywordImage Deblurring Non-parametric Pointwise P-norm Hyper-parameter
MOST Discipline Catalogue工学::计算机科学与技术(可授工学、理学学位)
Indexed ByEI
Funding ProjectNational Natural Science Foundation of China[61571438]
Citation statistics
Document Type会议论文
First Author AffilicationInstitute of Automation, Chinese Academy of Sciences
Recommended Citation
GB/T 7714
Binbin, Ma,Xiyuan, Hu,Silong, Peng. An adaptive pointwise P-norm regularization for image deconvolution[C],2019.
Files in This Item: Download All
File Name/Size DocType Version Access License
SP022.pdf(411KB)会议论文 开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Binbin, Ma]'s Articles
[Xiyuan, Hu]'s Articles
[Silong, Peng]'s Articles
Baidu academic
Similar articles in Baidu academic
[Binbin, Ma]'s Articles
[Xiyuan, Hu]'s Articles
[Silong, Peng]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Binbin, Ma]'s Articles
[Xiyuan, Hu]'s Articles
[Silong, Peng]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: SP022.pdf
Format: Adobe PDF
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.