基于特征点的高分辨率遥感影像城镇道路提取  

Urban Road Extraction from High Resolution Remote Sensing Image Based on Feature Points

作　　者：刘善磊[1] 赵银娣[2] 石善球[1] 朱绍东 张亮[1] 

机构地区：[1]江苏省基础地理信息中心,江苏南京210013 [2]中国矿业大学环境与测绘学院,江苏徐州221008 [3]山东方元地理信息工程有限责任公司,山东潍坊261000

出　　处：《测绘与空间地理信息》2012年第9期33-37,41,共6页Geomatics & Spatial Information Technology

基　　金：江苏省测绘科研项目(JSCHKY 201219);中国矿业大学科学研究基金项目(2007 B 011)资助

摘　　要：根据高分辨率遥感影像中城镇道路的特点,提出了一种基于特征点的道路信息提取方法。首先,对影像进行增强处理并选取感兴趣的子区域,利用改进的分水岭分割理论和阈值选择算法,结合八邻域检测方法得到道路的特征点;然后,利用回归分析方法在一定的坐标系统下得到每条道路的回归方程,根据端点坐标信息得到道路信息图;最后,利用数学形态学算法获取道路骨架图。结果表明,本方法能够精确有效地提取高分辨率遥感影像的城镇道路信息。In this paper, an approach of extracting urban road in the high resolution remote sensing image based on feature points is presented according to the characteristics of urban road. Firstly the road information in the image is enhanced and the interested sub - region is selected. Mathematic morphology algorithm is used. Secondly the road feature points are obtained based on improved water- shed segmentation algorithm, threshold select algorithm and the detection of eight neighborhoods. Then the regression equation of road is calculated using regression analysis in the definite reference frame, and the road information map is described based on the coordi- nate of endpoint. Finally the skeleton of road is done by morphological thinning and cropping. Experimental results show that the pro- posed method based on coordinates of feature points performs effectively on road extraction from high resolution remote sensing image- ry.

关 键 词：高分辨率遥感 分水岭分割 回归分析 数学形态学 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]