Mourad Talbi*, Mohamed Salim Bouhlel and Adnane Cherif Pages 484 - 494 ( 11 )
Background: In this paper, we propose a new image denoising technique which combines two denoising approaches.
Methods: The first one is a curvelet transform (cvt) based denoising technique. The second one is a two-stage image denoising by principal component analysis with local pixel grouping (LPG-PCA). This proposed technique consists at first step in applying the first approach to the noisy image in order to obtain a first estimate of the clean image. The second step consists in estimating the level of noise corrupting the original image. This estimation is performed by using a method of noise estimation from noisy images. The third step consists in using this first clean image estimation, the noisy image and this noise level estimate as inputs of the second image denoising system (LPGPCA based image denoising) in order to obtain the final denoised image.
Conclusion: The proposed image denoising technique was applied on a number of noisy images and the obtained results from PSNR and SSIM computations show its performance.
Curvelet transform, image denoising, PCA, local pixel grouping, noise estimation, noise corrupting.
Laboratoire de Nanomateriaux et Systemes des Energies Renouvelables (LaNSER) and Center of researches and technologies of energy of Borj Cedria; Univsersity of Sfax, Sfax, Sciences Electroniques, Technologie de l'Information et Telecommunications (SETIT), Sfax, Innov'Com Group, Signal Processing Laboratory, Sciences Faculty of Tunis, Tunis