A NOVEL INTER AND INTRA SCALE WAVELET DENOISING WITH ARBITRARY WINDOWS
Meena Srinivasan ¹ , S.Annadurai ^2
1 Dept of CSE, Government College of Technology, Coimbatore-641013, India.
² Government College of Engineering, Tirunelveli-627007, India.

Abstract:

This paper presents an efficient de-noising procedure for additive white Gaussian noise corrupted images, by exploiting both the inter and intra scale interdependencies that exist among the wavelet coefficients. The intra scale dependencies are better exploited by using arbitrary shaped windows with modified maximum likelihood variance estimation. By utilizing the persistence property of the wavelet coefficients, the Hidden Markov Tree (HMT) model captures the inter scale dependencies. Despite the simplicity in its concept and implementation, the proposed algorithm provides better PSNR and visual quality compared to the existing image denoising methods.

Keywords: Wavelet denoising, Gaussian noise removal, statistical modeling, inter intra scale dependency, arbitrary shaped windows.

BibTex:

@ARTICLE{P1150627002,

}

(Full Paper 498KB)