• REENA
• vinay kumar srivastava

Estimation of noise variance is an important parameter in many image processing applications such as denoising, compression, segmentation and edge detection. In this paper, literature survey on techniques of noise variance estimation has been presented and these techniques are compared. It has been observed that most of the estimation methods overestimate the noise variance for rich textured images like baboon, mountain, and grass etc. It is found that PCA based methods gives more accurate estimation in comparison to other methods for textured images. In this work, the PCA based method is applied for noise variance estimation in BayesShrink denoising. It competes well with the MAD based noise variance estimation method adopted in BayesShrink and outperforms in most of the experimental results as presented in this paper. The PCA based method provides a more accurate noise level estimation for low as well as high noise level for all types of images. Also in this work, different noise variance estimation methods are applied to BayesShrink denoising and the results are compared. The experimental result shows that the PCA based method improves the denoising performance of BayesShrink in contrast to other methods especially for textured images.

Gaussian white noise, BayesShrink, PCA, textured images, image denoising, and noise variance estimation

"Noise Variance Estimation of Textured Images Applied for BayesShrink Denoising", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 4, page no.578-585, April-2019, Available :http://www.jetir.org/papers/JETIR1904P90.pdf

"Noise Variance Estimation of Textured Images Applied for BayesShrink Denoising", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 4, page no. pp578-585, April-2019, Available at : http://www.jetir.org/papers/JETIR1904P90.pdf