摘要
给出了一种新的高分辨率合成孔径雷达(SAR)图像高速公路提取算法。首先对SAR图像进行FROST滤波,抑制斑点噪声,再使用OSTU算法对SAR图像进行二值化,然后使用腐蚀-膨胀算法与HOUGH变换对二值化后的SAR图像进行处理,提取出高速公路中的隔离带,再利用圆形模板匹配方法提取高速公路中单行道的道路中心线,最后基于单行道的宽度提取出整个高速公路。实验结果表明,该方法可以消除噪声与干扰的影响,完整地提取出高速公路。
A highway extraction method is proposed for high-resolution Synthetic Aperture Radar（SAR） image.Firstly,the speckle noise in SAR image is filtered out.Next,SAR image is processed by OSTU algorithm.Then,SAR image is processed by dilation-erosion algorithm and HOUGH transformation,and the isolation belt is extracted.After that,circular template matching is adopted to extract centerline of the one way.Finally,the whole highway is extracted based on the width of one way.Simulation results show that the proposed method can eliminate noises and interferences,and the complete highway can be extracted.
出处
《遥感信息》
CSCD
2012年第5期8-13,共6页
Remote Sensing Information
基金
2009年度中科院大科学装置维护改造项目
关键词
合成孔径雷达
道路提取
圆形模板匹配
Synthetic Aperture Radar（SAR）
road extraction
circular template matching
