Summary

Smart Info-Media Systems in Asia

2019

Session Number:RS3

Session:

Number:RS3-5

An Estimation of Standard Deviation of Gaussian Noise Using the Image Variation Components and the Denoising Filter

Takeshi Suzuki,  Hiroyuki Tsuji,  Tomoaki Kimura,  

pp.151-154

Publication Date:2019/9/4

Online ISSN:2188-5079

DOI:10.34385/proc.57.RS3-5

PDF download (847.5KB)

Summary:
There are methods based on PCA using iterative processing and MAD as methods to estimate the standard deviation σ of Gaussian noise in the Gaussian noise image. These methods can be estimated accurately if the Gaussian noise is large. However, in the case of low Gaussian noise with σ less than 10, its estimation accuracy is not good. In this paper, we extend the method based on denoising filter and MAD, and propose a estimation method of low Gaussian noise by using the difference. In this paper, it is clarified that the error can be reduced by 12% in the low Gaussian noise image compared with the conventional method using MAD.