新型快速中值滤波算法及在医学图像中的应用

Novel fast median filtering algorithm and its application in medical image

医学图像的处理具有数据量大、实时性强的特点。而传统中值滤波(SM)算法需对滤波窗口内所有的元素完全排序后确定中值,算法的时间复杂度高,执行效率低,无法满足医学图像实时处理的需要。为提高中值滤波算法的执行效率,提出一种新型的快速中值滤波算法,将基于排序查找中值的过程转换为基于分治查找中值的过程,有效地避免了对中值滤波而言不必要的完全排序运算,降低了算法的复杂度;同时在分治查找过程中,利用医学图像未受干扰时图像中像素值的变化是渐变的特性,优先选用中心点附近的像素值进行分治查找,达到提高算法效率的目的。理论分析和实验结果均验证了该算法的可行性和有效性。实验结果显示,该算法提高了医学图像的滤波速度,增强了医学图像处理的实时性。

Huge data and real time are the characteristics of medical image processing.But standard median filter need to sort all the pixels in the mask for finding the median value.So the algorithm is time consuming,and can’t meet the need of real time processing of medical image.To improve the efficiency of median filter,this paper proposed a novel fast median filter algorithm,which changed the process of finding median by sorting into dividing searching.In dividing searching,it used the pixels near middle of the mask for priority dividing searching.It could save the time for sorting.Theory analysis and simulation results prove the proposed algorithm is reasonable and effective.Experiment results show that the algorithm increases the speed of medical image filtering,enhances the real time processing of medical image