FAST TOTAL VARIATION BASED IMAGE RESTORATION UNDER MIXED POISSON-GAUSSIAN NOISE MODEL

Ghulyani, Manu and Arigovindan, Muthuvel (2018) FAST TOTAL VARIATION BASED IMAGE RESTORATION UNDER MIXED POISSON-GAUSSIAN NOISE MODEL. In: 15th IEEE International Symposium on Biomedical Imaging (ISBI), APR 04-07, 2018, Washington, DC, pp. 1264-1267.

PDF Iee_Int_Sym_Bio_Ima_1264_2018.pdf - Published Version Restricted to Registered users only Download (500kB) | Request a copy

Abstract

Image acquisition in many biomedical imaging modalities is corrupted by Poisson noise followed by additive Gaussian noise. MLE based restoration methods that use the exact Likelihood function for this mixed model with non-quadratic regularization are very few. While it has been demonstrated that total variation (TV) based regularization methods give better results, such methods that use exact Poisson-Gaussian Likelihood are slow. Here, we propose an ADMM based fast algorithm for image restoration using exact Poisson-Gaussian Likelihood function and TV regularization. Specifically, we propose a novel variable splitting approach that enables isolating the complexity in the exact MLE functional from the image blurring operation, allowing a fast Newton-like iteration on the MLE functional. This leads to a significantly improved convergence rate of the overall ADMM iteration. The effectiveness of the proposed method is demonstrated using restoration examples.

Item Type:	Conference Proceedings
Series.:	IEEE International Symposium on Biomedical Imaging
Publisher:	IEEE
Additional Information:	15th IEEE International Symposium on Biomedical Imaging (ISBI), Washington, DC, APR 04-07, 2018
Keywords:	Image Restoration; Maximum likelihood estimator (MLE); Alternating direction method of multipliers (ADMM); Poisson-Gaussian noise; Total variation; Regularization
Department/Centre:	Division of Electrical Sciences > Electrical Engineering
Date Deposited:	07 Feb 2019 08:58
Last Modified:	07 Feb 2019 08:58
URI:	http://eprints.iisc.ac.in/id/eprint/61671

Actions (login required)

View Item