基于边缘关键点检测线特征提取  

The Key Point Detection Based on Edge Line Feature Extraction

作　　者：范庆明[1] 王平[1] 高颖慧[1] 徐保勇[1] 费太勇[2] 

机构地区：[1]国防科学技术大学ATR实验室,湖南长沙410073 [2]空军雷达学院陆基预警监视装备系,湖北武汉430019

出　　处：《湖北民族学院学报（自然科学版）》2011年第4期377-381,共5页Journal of Hubei Minzu University(Natural Science Edition)

基　　金：国家自然科学基金项目(61103082)

摘　　要：为了从异源图像中提取目标线特性,针对传统的线特征提取存在漏警和虚警情况,提出边缘关键点拟合直线的方法,引入Helmholtz法则对直线提取进行约束.实验表明,该方法可以从异源图像中准确提取线特征,具有鲁棒性和实时性.In order to extract characteristics of the target line from the image of a different source,we propose the method of fitting straight line at the edge critical point,direction at leak alarm and false alarm in the traditional feature edge extraction.The Helmhotz law is introduced to bind the line extraction.Experiments show that the method can accurately extract the line characteristics from different source image,with robust and real-time characteristics.

关 键 词：线特征 边缘关键点 Helmholtz法则 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]