基于多个极化通道的SAR影像道路探测  

THE DETECTION OF ROADS IN SAR IMAGES BASED ON MULTI-POLARIMETRIC CHANNELS

作　　者：张广伟[1] 邓喀中[1] 张永红[2] 余海坤[2] 

机构地区：[1]中国矿业大学环境与测绘学院,徐州221008 [2]中国测绘科学研究院,北京100039

出　　处：《国土资源遥感》2006年第4期29-32,共4页Remote Sensing for Land & Resources

摘　　要：提出了一种从SAR影像的多个极化通道探测道路的新方法。首先,介绍了用边缘探测器构造道路探测器的方法;然后,将多参数统计假设检验的方法应用到全极化影像的信息融合中,并将该方法提取的结果与传统Tup in法提取的结果进行了比较。实验表明,该方法对全极化SAR影像道路边缘的提取效果较好,具有一定的适用性。A new method for detection of roads in multi - channel polarimetric SAR images is proposed in this paper. The technology for construction of the road detector from edge detectors was described, the approach based on multivariate statistical hypothesis tests was applied to information fusion of SAR images, and the extracting result from two traditional Tupin methods and that from the method proposed in this paper were compared with each other. Experiments show that the method proposed in this paper, when applied to detecting roads on a full - polarimetric SAR image, can yield relatively satisfactory result.

关 键 词：SAR影像 道路 边缘探测 多通道 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]