海岸线是海洋与陆地的交界线,全球气候变化将引起海岸线的变迁,极地区域拥有最大的冰川,冰川的消融将导致全球海岸线的迅速变迁,因此研究极地海岸线具有重要的科学意义。由于潮汐等影响,卫星影像提取的海洋与陆地的交界线只是海岸水边线,本文以格陵兰岛海岸为实验对象,采用Gamma滤波方法对合成孔径雷达(SAR)数据的相干斑进行预处理,提出了应用支持向量机(SVM)和Canny边缘检测算子的海岸水边线提取方法,基于TerraSAR-X影像的实验结果表明,本文提取的格陵兰海岸水边线准确率都达60%以上。%Global climate change will cause changes in coastlines,the boundary between land and sea.Polar regions have the largest glaciers,and melting of these glaciers could lead to rapid changes in coastlines throughout the world.Therefore,research on polar coastlines is important.Satellite images can be used to determine the coastal water boundary line.In this paper,we investigated the coast of Greenland using the Gamma filtering method for speckle preprocessing of Synthetic Aperture Radar (SAR)data.We present a method combining the Support Vec-tor Machine (SVM)and the Canny edge detection operator to extract the coastal water boundary line.Using this approach,the extraction accuracy of the Greenland coastline from TerraSAR-X images was up to 60 percent.
展开▼