Image Denoising Based On Wavelet Transform And Median Filter

Wavelet transform is a populous signal processing method in recent years. Wavelets have advantages of local analysis and multiresolution analysis, and they are very suitable for image processing. There is often noise in real images. In order to clear off the noise in image and meanwhile preserve the image edge information, we propose in the paper a efficient technique based on wavelet transform and median filter. Based on the analysis of Gaussian and impulse noise in wavelet transform, our method filters the high-frequency subband images combined to the property of median filter, then thresholds the subband images in wavelet-domain. The experiment results show that the method not only efficiently remove mixed Gaussian and impulse noise in image, but also preserve the image edge information. The filtering performance is better than de-noising by Soft-Thresholding.