Robust noise estimation based on noise injection

2Citations
Citations of this article
2Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Noise estimation is an important premise for image denoising and the related research therefore has drawn increasing attention and interest. Recent studies show that the distribution mode of local variances in natural image can be used as a simple yet efficacious estimator of the additive noise variance, no matter what distribution the noise follows. However, this type of method has the limitation that the target image must have a sufficiently large area with low pixel value variations. Furthermore, this type of noise estimator almost always lead to overestimation without taking into account the mode of local variance distribution of the noise-free image in textural regions. To improve the accuracy of distribution-mode analysis type of noise estimation and to resolve the problem of overestimation, we propose a novel algorithm using a cascade of wavelet sub-band estimation and noise-injection based rectification. The proposed algorithm reduces the detrimental influence of textural image area, and therefore alleviating overestimation of the noise variance. Extensive experiments and comparative study show the reliability and superiority the proposed method over some existing competitors. © 2012 Springer-Verlag.

Cite

CITATION STYLE

APA

Tang, C., Yang, X., & Zhai, G. (2012). Robust noise estimation based on noise injection. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7674 LNCS, pp. 142–152). https://doi.org/10.1007/978-3-642-34778-8_13

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free