计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (16): 216-218.

• 工程与应用 • 上一篇    下一篇

基于区域生长的道路和桥梁识别方法的研究

李桂芹,尹 东,薛晨荣   

  1. 中国科学技术大学 信息处理中心,合肥 230027
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-06-01 发布日期:2007-06-01
  • 通讯作者: 李桂芹

Study of road and bridge recognition based on region growing

LI Gui-qin,YIN Dong,XUE Chen-rong   

  1. Information Processing Center of University of Science and Technology,Hefei 230027,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-06-01 Published:2007-06-01
  • Contact: LI Gui-qin

摘要: 针对高分辨率可见光图像中的道路和桥梁目标,提出了一种基于区域生长的道路和桥梁自动识别的方法。利用区域生长对原始图像进行粗分类,有效去除利用Canny算子进行边缘检测后得到的“弱边缘”,然后利用规则去除无效线段,再检测平行线,对不连续的线段进行连接,最后根据目标模型的特征标识出道路和桥梁目标。实验结果表明,在不需要人工干预的条件下,该方法能够快速有效地自动识别出道路和桥梁目标。

Abstract: In this paper,a method to solve the problem of road and bridge recognition based on region growing from high resolution optical images is presented.First,in order to remove “thin edges” which are produced by edge detection with Canny operator,we use region growing to classify original image approximately,then use rules to remove useless segments,detect parallel and connect discrete segments,and at last identify road and bridge according to target model.Experimental result shows that road and bridge can be recognized fast and efficiently without manual work.