Medical Image Denoising Based on Wavelet Transform and Median Filtering
YE Hongjin, Xueying Zhang · Journal of Taiyuan University of Technology · 2005
Wavelet transform is a kind math tool developed in recent decade. It has better property than Fourier transform. Wavelet transform is widely used in signal process. This paper introduces the definition of discrete wavelet transforms the principle of two dimensions wavelet transform decomposition and reconstruction then median filtering. It describes a new method for denoising of medical CT images. Mixed Gaussian and Impulse noise can be reduced by using wavelet transform then median filtering. Experimental results show the method can improve the image quality and has good visual results. Its effect is better than pure wavelet transform,pure median filtering and median filtering then wavelet transform denoising. It is an ideal denoising algorithm for CT images including mixed Gaussian and Impulse noise.