基于球形结构元素的两种三维形态滤波方法  

Two 3D Morphological Filtering Methods Based on Sphere Structuring Element

作　　者：张彬[1,2] 毛先成[1,2] 

机构地区：[1]中南大学有色金属成矿预测教育部重点实验室,湖南长沙 [2]中南大学地球科学与信息物理学院,湖南长沙

出　　处：《图像与信号处理》2016年第1期33-42,共10页Journal of Image and Signal Processing

基　　金：国家自然科学基金资助项目(41472301,41172297)。

摘　　要：基于数学形态学的三维形态滤波在采用大规模结构元素进行运算时,变得异常耗时,若采用的结构元素为球形,则可以使用欧式距离变换代替基本的形态学运算。以某地质体二值图像为背景,对两种三维形态滤波方法进行对比分析,实验结果表明,基于球形结构元素的两种形态滤波方法的滤波效果完全相同,但基于欧式距离变换的形态滤波方法在针对大规模球形结构元素时更为有效。3D morphological filtering based on mathematical morphology can become very time-consuming when using a large-scale structuring element involving the operation;however, if the structuring element is sphere, the Euclidean distance transform (EDT) could be used to perform basic morphological operations instead. Taking some geologic body as a case study, the comparison analysis results show that two filtering methods based on sphere structuring element have the same effects, whereas, the method based on EDT is more effective and applicable for large-scale structuring element involved.

关 键 词：三维形态滤波 球形结构元素 数学形态学 欧式距离变换 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]