Analysis And Comparison Of Several De-noising Methods For Medical Image

Noises are inevitably introduced to medical images because of various factors in medical imaging. The noises in medical images degrade the quality of images, blurring the boundaries and suppressing structural details, thus bring difficulties to medical diagnosis. Therefore, the key to medical image de-noising is to remove the noises while preserving important features. Image de-noising methods can be categorized into spatial de-noising and transform domain de-noising. Partial differential equation based methods and bilateral filtering are two representative techniques, wavelet de-noising and sparse representation based de-noising are two kinds of popular transform domain algorithms. In this paper, the partial differential equation based methods including the classical P-M diffusion model and the total variation (TV) model, bilateral filtering, wavelet de-noising methods are first introduced. We analyzed and compared the peak signal to noise ratio (PSNR) and the difference image of the image de-noised by the five de-noising algorithms and the de-noising time of the five de-noising algorithms to provide convenience for targeted choosing of de-noising methods.