A Convex Variational Approach for Image Deblurring With Multiplicative Structured Noise

The restoration of images corrupted by blurring and structured noise has attracted growing attention in the domains of image processing and computer vision. However, many works only focus on the restoration of the images degraded by blurring and additive structured noise or multiplicative structured...

Full description

Saved in:
Bibliographic Details
Published in:IEEE access 2020, Vol.8, p.37790-37807
Main Authors: Wu, Tingting, Li, Wei, Li, Lihua, Zeng, Tieyong
Format: Article
Language:English
Subjects:
Additives
ADMM
Blurring
Computer vision
convex variational model
Degradation
Image deblurring
Image processing
Image restoration
Manganese
Mathematical model
Multiplication
multiplicative noise
Noise
Noise reduction
Numerical models
structured noise
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The restoration of images corrupted by blurring and structured noise has attracted growing attention in the domains of image processing and computer vision. However, many works only focus on the restoration of the images degraded by blurring and additive structured noise or multiplicative structured noise separately. It is still a challenge and an open problem to restore degraded images with blurring and multiplicative structured noise, simultaneously. In this paper, based on the total variation (TV), the statistical property of the Gamma noise and the maximum a posteriori (MAP) estimator, we obtain a convex variational model to recover blurred images with multiplicative structured noise. Especially, to get this convex model, we reformulate the prior assumption of the images degradation model by division instead of multiplication. For solving this convex model, an effective alternating direction method of multipliers (ADMM) is employed. Numerical experiments are presented to illustrate the effectiveness and efficiency of the proposed model.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2020.2974913