期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B7
页码：125-130
出版社：Copernicus Publications
摘要：The uncertain factors of automatic extracting roads from the Synthetic Aperture Radar (SAR) imagery have been considered in details in the paper. We consider that the key point of the extraction is the accurate detection of the road points in the image. So, to get the good result, at first, through analyzing the roads features in the SAR image, we got the local model of roads structure. Then utilizing the statistic characteristic of the SAR image, we obtained the ratio probability density function (pdf) of the adjacent region to the road region. As follow, we analyzed the uncertainty of the parameters in the pdf. So we got the adaptable rules about setting the parameters. Subsequently, we used the K-S hypothesis testing to get the roads points in the image. Finally, we studied various objective quality measures to evaluate the extraction results. The practices have shown that the results are in agreement with the theoretical analysis, and so men can enhance the performance of the extraction through setting adaptable parameters
