基于小波变换和中值滤波的医学图像去噪  

Medical Image Denoising Based on Wavelet Transform and Median Filtering

作　　者：叶鸿瑾[1] 张雪英[1] 何小刚[1] 

机构地区：[1]太原理工大学信息工程学院

出　　处：《太原理工大学学报》2005年第5期511-514,共4页Journal of Taiyuan University of Technology

基　　金：国家自然科学基金资助项目(60472094);教育部留学回国科研启动基金资助项目(教外司留[2004]176号文件);山西省留学回国人员科研基金资助项目(200224)

摘　　要：简单介绍了离散小波变换、二维小波变换分解与重构和中值滤波的原理,提出了利用小波变换、中值滤波对含有高斯和脉冲两者混合噪声的医学CT图像进行去噪的一种新方法。实验结果表明:这种方法能够有效改善图像质量,较好地保持图像视觉效果,降低图像噪声;此方法的效果优于单纯的小波变换或单纯的中值滤波或先中值滤波再小波变换去噪的方法,是去除CT图像中含高斯和脉冲两者混合噪声的一种比较理想的方法。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.

关 键 词：小波变换 中值滤波 医学图像 图像去噪 

分 类 号：TN911.73[电子电信—通信与信息系统]