New Image Denoising Algorithm Based on Improved Grey Prediction Model

Eliminating the noise and protecting the details in the image are the purpose of the noise reduction. The theory based on grey prediction model introduces a nonlinear filter, which is used to improve the quality of the image denoising and meanwhile keep the image details. The basic theory and the method of grey prediction model are introduced. The improved algorithm can detect and obtain more precise edge pixels without noise. The experiment results and the analysis of Signal-to-Noise show that when the noise property reaches to 40%, the PSNR with the new algorithm is always better than the original algorithm and conventional median filtering. The successful algorithm indicates that it is feasible and effective to use grey prediction model to process the image.