一种自适应双阈值中值滤波方法  

Adaptive Dual-threshold Median Filtering Method

作　　者：宗永胜 胡晓辉[1] 张荣光[1] 

机构地区：[1]兰州交通大学电子与信息工程学院,兰州730070

出　　处：《小型微型计算机系统》2017年第7期1642-1647,共6页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61163009)资助;甘肃省科技支撑计划项目(144NKCA040)资助

摘　　要：为解决中值滤波算法在图像去噪及保护图像细节方面的不足,提出一种自适应双阈值中值滤波方法.在该方法中,首先采用基于均值的双阈值选取方法对脉冲噪声进行判别,然后在噪声滤除阶段使用滤波窗口中像素的中值或其邻域中已经经过滤波处理后的像素的中值代替噪声像素,对于非噪声像素则保持其像素值不变.因为阈值的选取是随滤波窗口的滑动而变化的,这就体现了阈值选取的自适应性.实验结果表明,该方法能有效的滤除脉冲噪声,同时也可以更好的保持图像的边缘及细节信息,通过定量和定性分析,本方法的滤波性能要优于中值滤波及其他改进方法.To deal with the deficiency in terms of denoising and preserving image details of the standard median filter algorithm,an adaptive dual-threshold median f＇dtering method is proposed. In the method,the impulse noise detection is based on dual-threshold estimated by the mean-based strategy. The noisy pixels are filtered through the replacement of the median value of the filtering window or the previously restored neighboring adjacent noisy pixels of the window, and the values of the noise-free pixels are left unchanged. Threshold＇s selection is changed along with the sliding of the filtering window ,so it justifies the adaptive operation of the method. Experimental results justify that the proposed method can not only denoise effectively but also preserve the edge and details of image desirablly. Both quantitative and qualitative analysis shows the superiority of the method over other filtering algorithms.

关 键 词：中值滤波 自适应双阈值 均值 边缘信息 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]