Color Image Restoration Method for Gaussian Noise Removal

Abstract

A new approach to the restoration of color images corrupted by Gaussian noise is presented. The proposed technique adopts a multipass processing approach that gradually reduces the noise in the color information components of the image. Two different models for data smoothing are proposed based on the different classes of noisy pixels. The subsequent algorithm for edge detection is designed to better appraise the noise cancellation behavior of our filter from the point of view of human perception. This method does not require any "a priori" knowledge about the amount of noise corruption. Experimental results show that the filtering performance of the proposed approach is very satisfactory and accurate edge maps are achieved even in the presence of highly corrupted data. © Springer-Verlag Berlin Heidelberg 2010.