基于高分遥感的金沙江流域滑坡识别——以巴塘县王大龙村为例

金沙江流域因两岸地势陡峭、软弱岩层发育、降雨集中等，使得流域内滑坡灾害分布密集。高分辨率遥感是滑坡识别的重要手段，但通过目视解译法开展的大范围滑坡灾害识别，具有工作量大、效率低的特点。针对此问题，本文采用基于面向对象的分类方法，提出了利用滑坡灾害的光谱、形状、空间等特征进行区域内滑坡灾害的快速识别。同时，选取金沙江流域巴塘县王大龙村区段进行了滑坡识别提取试验，区域内利用面向对象分类方法识别出滑坡18处，其中12处与目视解译结果相同，一致性为75%；发现3处目视解译未识别出的隐蔽性滑坡。结果表明，该方法识别效果较好，可为后续的金沙江流域乃至川藏铁路沿线的大范围滑坡识别提取及滑坡编目工作提供参考。

Landslide identification in Jinsha River basin based on high-resolution remote sensing:taking Wangdalong village of Batang county as an example

Because of the steep terrain,the development of soft rock,and the concentration of rainfall on both sides of the Jinsha River basin,the landslide disasters are densely distributed in the basin.High-resolution remote sensing is an important means of landslide identification,but the visual interpretation method for large-scale landslide hazard identification has the characteristics of heavy workload and low efficiency.In this paper,the object-oriented classification method is used to identify the large-scale landslides in Jinsha River basin.The spectrum,shape and space characteristics of landslides are used to identify the landslides in the region.At the same time,the Wangdalong village section of Batang county in the Jinsha River basin is selected for the identification and extraction of landslides.In this area,18 landslides are identified by object-oriented classification method,12 of which are the same as the results of visual interpretation,and the consistency is 75%.In addition,3 hidden landslides are found which are not identified by visual interpretation.The results show that the recognition effect of this method is good,which can provide a reference for the subsequent large-scale landslide identification extraction and landslide cataloging work in the Jinsha River basin and even along the Sichuan-Tibet railway.