一种基于数学形态学的遥感图象边缘检测算法  

A Mathematical Morphological Algorithm for Edge Detection in Re motely Sensed Image

作　　者：王宇[1] 王乘[1] 刘吉平[1] 

机构地区：[1]华中科技大学,武汉430074

出　　处：《计算机工程与应用》2003年第30期91-93,共3页Computer Engineering and Applications

摘　　要：为有效提取边缘密度大的遥感图象中的边缘信息,该文提出了一种利用灰度形态变换原理进行检测的有效算法,首先利用开-闭形态滤波器降低输入图象的噪声,然后将二值图象的边缘提取算法推广到灰度图象中加以应用。实验结果证明,此方法优于灰度形态梯度法,同时也优于Canny算子。In this paper,we introduce a mathematical morphological algorithm for edge detection in remotely sensed im-ages of high density edges.Firstly,we filter the noise by using open-closing morphology filter,then use an edge detec-tion algorithm derived from binary mathematical morphology.It has been demonstrated that this algorithm do more effec-tively than Morphological gradient algorithm and Canny algorithm.

关 键 词：边缘检测 数学形态学 形态梯度 

分 类 号：TP317.3[自动化与计算机技术—计算机软件与理论]