Edge preserving image denoising using convolutions and Grey Forecast Model

Most of filtering algorithms can remove noise to a large extent, but make the image edge blurred and distorted as a result of implementing uniformly across the image and thus tending to modify both noise and noise-free pixels. When an image is covered with high density noise many original information of the image can't be seen and known. That means the image could be taken as a grey system. This paper proposes an algorithm for both image denoising and image edge preserving. The proposed method is based on noise recognition and grey theory. It consists of a novel model for noise removal using grey model.. This method provides only "noise" pixels for processing. Computer simulations are carried out to assess the performance of the proposed method using test images which are corrupted by 20% and 50% salt & pepper noise, respectively. Experimental results are given to demonstrate the denoising performance of our algorithm. It gives better result in visual quality and numerical results in terms of both the mean square error and peak signal-to-noise ratio (PSNR) when compared to three other existing methods. The experimental results show the great potential of the proposed model.