• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2010, Vol. 32 ›› Issue (8): 71-74.doi: 10.3969/j.issn.1007130X.2010.

• 论文 • 上一篇    下一篇

基于方向ROEWA算子的高分辨率SAR图像道路提取

吴禹昊,陈天泽,粟毅   

  1. (国防科学技术大学电子科学与工程学院,湖南 长沙 410073)
  • 收稿日期:2009-01-12 修回日期:2009-04-09 出版日期:2010-07-25 发布日期:2010-07-28
  • 作者简介:吴禹昊(1983),男,辽宁沈阳人,硕士生,研究方向为遥感信息处理;陈天泽,讲师,研究方向为遥感信息处理与3S集成;粟毅,教授,研究方向为超宽带雷达系统、信号与信息处理、遥感信息处理。

Road Extraction from High Resolution SAR Images Based on the Directional ROEWA Detector

WU Yuhao,CHEN Tianze,SU Yi   

  1. (School of Electronics Science and Engineering,National University of Defense Technology,Changsha 410073,China)
  • Received:2009-01-12 Revised:2009-04-09 Online:2010-07-25 Published:2010-07-28

摘要:

指数加权均值比率ROEWA算子是SAR图像阶跃边缘提取中较为常用的算法,但该方法只能提取边缘强度,不能得到边缘方向。本文在ROEWA算子的基础上,利用添加方向模板的方法,提出了获取边缘方向的ROEWA改进算法。并根据得到的边缘方向,分别提出了一种边缘筛选算法和快速的Hough变换算法,从而提取出图像中的直线信息。实验表明,本文方法准确、高效,且对SAR图像中的相干斑噪声具有很好的抑制作用。

关键词: SAR图像, ROEWA算子, 边缘方向, Hough变换, 道路提取

Abstract:

The ROEWA (Ratio Of Exponentially Weighted Averages) edge detector has been widely used in the extraction of the step edges in SAR images. The detector can only get the edge intensity, but it can not get the edge direction. A modified ROEWA edge detector is proposed to gain the edge direction with directional templates. After edge detection, a new edge selecting algorithm and a fast Hough transformation are proposed based on the edge direction, which can gain the straight lines in the images. The simulated result experiments the accuracy and the efficiency of the algorithm proposed, and the speckle noise is suppressed.

Key words: SAR image;ROEWA detector;edge direction;Hough transformation;road extraction