Poisson Noise Reduction with Higher-Order Natural Image Prior Model
Feng, Wensen1; Qiao, Hong2; Chen, Yunjin3
Source PublicationSIAM JOURNAL ON IMAGING SCIENCES
2016
Volume9Issue:3Pages:1502-1524
SubtypeArticle
AbstractPoisson denoising is an essential issue for various imaging applications, such as night vision, medical imaging, and microscopy. State-of-the-art approaches are clearly dominated by patch-based non local methods in recent years. In this paper, we aim to propose a local Poisson denoising model with both structural simplicity and good performance. To this end, we consider a variational modeling to integrate the so-called fields of experts (FoE) image prior, that has proven an effective higher-order Markov random fields model for many classic image restoration problems. We exploit several feasible variational variants for this task. We start with a direct modeling in the original image domain by taking into account the Poisson noise statistics, which performs generally well for the cases of high signal-to-noise ratio (SNR). However, this strategy encounters problem in cases of low SNR. Then we turn to an alternative modeling strategy by using the Anscombe transform and Gaussian statistics derived data term. We retrain the FoE prior model directly in the transform domain. With the newly trained FoE model, we end up with a local variational model providing strongly competitive results against state-of-the-art nonlocal approaches, meanwhile bearing the property of simple structure. Furthermore, our proposed model comes along with an additional advantage, that the inference is very efficient as it is well suited for parallel computation on GPUs. For images of size 512 x 512, our GPU implementation takes less than 1 second to produce state-of-the-art Poisson denoising performance.
KeywordPoisson Denoising Fields Of Experts Anscombe Root Transformation Nonconvex Optimization
WOS HeadingsScience & Technology ; Technology ; Physical Sciences
DOI10.1137/16M1072930
WOS KeywordINVERSE PROBLEMS ; OPTIMIZATION ; SPARSITY ; RESTORATION ; INTENSITY ; INFERENCE ; MRFS
Indexed BySCI
Language英语
WOS Research AreaComputer Science ; Mathematics ; Imaging Science & Photographic Technology
WOS SubjectComputer Science, Artificial Intelligence ; Computer Science, Software Engineering ; Mathematics, Applied ; Imaging Science & Photographic Technology
WOS IDWOS:000385277200024
Citation statistics
Cited Times:1[WOS]   [WOS Record]     [Related Records in WOS]
Document Type期刊论文
Identifierhttp://ir.ia.ac.cn/handle/173211/12572
Collection复杂系统管理与控制国家重点实验室_机器人理论与应用
Corresponding AuthorChen, Yunjin
Affiliation1.Univ Sci & Technol Beijing, Sch Automat & Elect Engn, Beijing 100083, Peoples R China
2.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China
3.Graz Univ Technol, Inst Comp Graph & Vis, A-8010 Graz, Austria
Recommended Citation
GB/T 7714
Feng, Wensen,Qiao, Hong,Chen, Yunjin. Poisson Noise Reduction with Higher-Order Natural Image Prior Model[J]. SIAM JOURNAL ON IMAGING SCIENCES,2016,9(3):1502-1524.
APA Feng, Wensen,Qiao, Hong,&Chen, Yunjin.(2016).Poisson Noise Reduction with Higher-Order Natural Image Prior Model.SIAM JOURNAL ON IMAGING SCIENCES,9(3),1502-1524.
MLA Feng, Wensen,et al."Poisson Noise Reduction with Higher-Order Natural Image Prior Model".SIAM JOURNAL ON IMAGING SCIENCES 9.3(2016):1502-1524.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Feng, Wensen]'s Articles
[Qiao, Hong]'s Articles
[Chen, Yunjin]'s Articles
Baidu academic
Similar articles in Baidu academic
[Feng, Wensen]'s Articles
[Qiao, Hong]'s Articles
[Chen, Yunjin]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Feng, Wensen]'s Articles
[Qiao, Hong]'s Articles
[Chen, Yunjin]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

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