Semi-blind image deblurring by a proximal alternating minimization method with convergence guarantees

•Our method successfully solves a nonconvex semi-blind image restoration model.•Our method is validated, both theoretically and numerically.•The proposed algorithm is efficient and stable.•Any sequences generated by our algorithm converge to a local minimized point. The semi-blind image deblurring p...

Full description

Saved in:
Bibliographic Details
Published inApplied mathematics and computation Vol. 377; p. 125168
Main Authors Dou, Hong-Xia, Huang, Ting-Zhu, Zhao, Xi-Le, Huang, Jie, Liu, Jun
Format Journal Article
LanguageEnglish
Published Elsevier Inc 15.07.2020
Subjects
Convergence analysis
Nonconvex optimization problem
Proximal method
Semi-blind image deblurring
Proximal method
Semi-blind image deblurring
Nonconvex optimization problem
Convergence analysis
Online AccessGet full text

Cover

More Information
Summary:•Our method successfully solves a nonconvex semi-blind image restoration model.•Our method is validated, both theoretically and numerically.•The proposed algorithm is efficient and stable.•Any sequences generated by our algorithm converge to a local minimized point. The semi-blind image deblurring problem aims to simultaneously estimate the clean image and the point spread function (PSF), which results in a (jointly) nonconvex optimization problem. In this paper, we develop an efficient algorithm to tackle the corresponding minimization problem based on the framework of the proximal alternating minimization (PAM). We also establish the convergence of the proposed algorithm under a mild assumption. Numerical experiments demonstrate our approach could obtain a more robust performance than the related state-of-the-art semi-blind image deblurring method.
ISSN:0096-3003
1873-5649
DOI:10.1016/j.amc.2020.125168