Email this article A Nonlocal Image Denoising Algorithm Using the Structural Similarity Metric
- Authors: Dovganich A.A.1, Krylov A.S.1
-
Affiliations:
- Faculty of Computational Mathematics and Cybernetics, Moscow State University
- Issue: Vol 45, No 4 (2019)
- Pages: 141-146
- Section: Article
- URL: https://journals.rcsi.science/0361-7688/article/view/176804
- DOI: https://doi.org/10.1134/S0361768819040029
- ID: 176804
Cite item
Open Access
Access granted
Subscription Access
Abstract
A new image denoising algorithm is proposed. It is a version of the nonlocal means (NLM) algorithm and uses a metric based on the CMCS modification of the structural similarity index (SSIM). The potentials of this metric for constructing the weighting function in the NLM method using the decomposition of this metric into components and specifying a physically justified weighting function for each component are demonstrated. The results produced by the modified method are compared with the results produced by the basic NLM algorithm, which uses the metrics L2 and SSIM for calculating the metric weights.
About the authors
A. A. Dovganich
Faculty of Computational Mathematics and Cybernetics, Moscow State University
Email: kryl@cs.msu.ru
Russian Federation, Moscow, 119991
A. S. Krylov
Faculty of Computational Mathematics and Cybernetics, Moscow State University
Author for correspondence.
Email: kryl@cs.msu.ru
Russian Federation, Moscow, 119991
