Open Access
Translator Disclaimer
2017 Total variation based denoising methods for speckle noise images
Arundhati Bagchi Misra, Ethan Lockhart, Hyeona Lim
Involve 10(2): 327-344 (2017). DOI: 10.2140/involve.2017.10.327

Abstract

In this paper, we introduce a new algorithm based on total variation for denoising speckle noise images. Total variation was introduced by Rudin, Osher, and Fatemi in 1992 for regularizing images. Chambolle proposed a faster algorithm based on the duality of convex functions for minimizing the total variation, but his algorithm was built for Gaussian noise removal. Unlike Gaussian noise, which is additive, speckle noise is multiplicative. We modify the original Chambolle algorithm for speckle noise images using the first noise equation for speckle denoising, proposed by Krissian, Kikinis, Westin and Vosburgh in 2005. We apply the Chambolle algorithm to the Krissian et al. speckle denoising model to develop a faster algorithm for speckle noise images.

Citation

Download Citation

Arundhati Bagchi Misra. Ethan Lockhart. Hyeona Lim. "Total variation based denoising methods for speckle noise images." Involve 10 (2) 327 - 344, 2017. https://doi.org/10.2140/involve.2017.10.327

Information

Received: 19 December 2015; Revised: 26 February 2016; Accepted: 19 March 2016; Published: 2017
First available in Project Euclid: 13 December 2017

zbMATH: 1366.94046
MathSciNet: MR3574304
Digital Object Identifier: 10.2140/involve.2017.10.327

Subjects:
Primary: 68U10 , 94A08
Secondary: 49K20 , 65K10 , 65M06 , 65N06

Keywords: anisotropic diffusion , Chambolle algorithm , Denoising , fast speckle denoising , speckle noise , total variation (TV) model

Rights: Copyright © 2017 Mathematical Sciences Publishers

JOURNAL ARTICLE
18 PAGES


SHARE
Vol.10 • No. 2 • 2017
MSP
Back to Top