Reaction-Diffusion Equation Based Image Denoising Algorithm

A Reaction-diffusion equation based image denoising algorithm (RDEID) is proposed for the removal of noises in corrupted images. By exploiting reactiondiffusion equation, the isotropic diffusion is performed to noisy pixels, the processed image is quantized and denoised. As a result, the details of image are effectively preserved during removing noises. Extensive experimental results show that the proposed algorithm has better performance in terms of evaluation metrics, such as Peak signal-to-noise ratio (PSNR), Image fidelity (IF) and Mean square error of noise suppression (MSENS).