基于模糊聚类的SAR图像道路检测  

Road Detection in SAR Images Based on Fuzzy Clustering

作　　者：赵晖[1] 孙进平[1] 王文光[1] 毛士艺[1] 

机构地区：[1]北京航空航天大学电子信息工程学院,北京100191

出　　处：《遥测遥控》2009年第2期34-39,共6页Journal of Telemetry,Tracking and Command

摘　　要：提出一种基于模糊聚类的SAR图像道路检测算法。首先根据SAR图像中道路的特点,使用FCM(模糊C均值)方法将图像进行模糊分类,获得大致的道路区域;然后通过重叠窗口的Hough变换,得到候选道路段;最后,根据候选道路之间的线性度、邻近度,以及道路所在位置图像的分类情况,对道路进行最终连接与筛选,获得检测结果。仿真结果表明,算法具有对道路定位准确、鲁棒性强的特点。A road detection method in SAR images based on fuzzy clustering is proposed in this paper.Firstly the fuzzy C mean(FCM) method is used to classify the SAR image,and the road pixels are extracted.Then the image is divided into sev-eral blocks,and the Hough transform is used to extract the segment candidates in each block.And then the proximity,the cocurvilinearity and the membership of the segment candidates are calculated to connect the roads.Finally,the false road is removed.The results of experiments indicate the availability and robust of the method.

关 键 词：道路检测 模糊C均值 HOUGH变换 合成孔径雷达图像 

分 类 号：TN957.52[电子电信—信号与信息处理] TP274.4[电子电信—信息与通信工程]