一种改进的基于柔性形态学的图像边缘检测方法  

An Improved Image Edge Detection Method Based on Adaptive Soft Mathematical Morphology

作　　者：李积英[1] 

机构地区：[1]兰州交通大学,甘肃兰州730070

出　　处：《平顶山学院学报》2011年第5期75-78,共4页Journal of Pingdingshan University

基　　金：国家自然科学基金(60962004);国家863高技术研究发展计划基金(2006AA02Z499)

摘　　要：在介绍柔性形态学的基础上,提出先利用迭代算法得到图像分割的最佳阈值,以增强图像的边缘.再利用多尺度柔性边缘检测算子得到图像的边缘,该方法根据不同尺度边缘图像所含信息量的多少确定边缘的合成,所以能实现图像的自适应边缘检测.实验结果证明:该方法与传统边缘检测算子提取的结果比较,可以更好地抑制噪声,得到的边缘在连续性和平滑方面均得到了提高.Based on the soft mathematical morphology,this paper first proposes using iterative algorithm for optimal threshold of image segmentation in order to enhance image edge,then using Multi-scale soft mathematical edge detection operator for image edge.According to the amount of information contained in edge images in different yardsticks,this method determines the synthesis of the edge,so as to achieve adaptive edge detection of images.Experimental results show that compared with the results extracted from the traditional edge detection operators,this method can suppress noise better.The edges that have been gained are improved in terms of continuity and smoothness.

关 键 词：数学形态学 图像边缘 柔性形态学 迭代算法 噪声 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]