Abstract
In this paper a new combination of image priors is introduced and applied to Bayesian image restoration. Total Variation (TV) image prior preserves edge structure while imposing smoothness on the solutions. However, it does not perform well in textured areas. To alleviate this problem we propose to combine TV with the Poisson Singular Integral (PSI) image prior, which is able to preserve image textures. The proposed method utilizes a bound for the TV image model based on the majorization-minimization principle, and performs maximum a posteriori Bayesian inference. In the experimental section the proposed approach is tested on synthetically degraded images with different levels of spatial activity and areas with different types of texture. Since the proposed method depends on a set of parameters, an analysis, about their impact on the final restorations, is carried out.
Original language | English (US) |
---|---|
Title of host publication | 2013 Proceedings of the 21st European Signal Processing Conference, EUSIPCO 2013 |
Publisher | European Signal Processing Conference, EUSIPCO |
ISBN (Print) | 9780992862602 |
State | Published - Jan 1 2013 |
Event | 2013 21st European Signal Processing Conference, EUSIPCO 2013 - Marrakech, Morocco Duration: Sep 9 2013 → Sep 13 2013 |
Other
Other | 2013 21st European Signal Processing Conference, EUSIPCO 2013 |
---|---|
Country | Morocco |
City | Marrakech |
Period | 9/9/13 → 9/13/13 |
Keywords
- Bayesian image restoration
- Deblurring
- Poisson Singular Integral
- Total Variation
ASJC Scopus subject areas
- Signal Processing
- Electrical and Electronic Engineering