【摘要】 针对两河口水电站库岸山高坡陡、面积广等特点，本文在收集区域地质、地形等资料的基础上，提出结合光学遥感的合成孔径雷达干涉测量(InSAR)技术滑坡识别方法，识别库岸滑坡隐患信息，分析滑坡分布特征。同时，考虑侧视雷达影像在高山峡谷区域几何畸变问题，采用改进R-index可视性分析模型，提取可视性较好区域进行InSAR数据处理，综合D-InSAR处理结果与光学遥感解译结果识别滑坡隐患。结果表明，本文选取的侧视雷达影像可提取48.7%可视区域。共识别滑坡29处，其中，通过D-InSAR技术识别滑坡8处，通过光学遥感技术识别滑坡28处，重合7处，2种技术识别结果一致性较高，互补作用明显。两河口水电站库岸滑坡大多分布在30°～45°的斜坡面上，并且植被覆盖少的凹岸或脊状山体处更易滑坡发育。更多还原

【Abstract】 According to the characteristics of Lianghekou Hydropower Station reservoir bank, such as high slope and wide area, on the basis of collecting regional geological and topographic data, this paper puts forward a landslide identification method based on interferometric synthetic aperture radar(InSAR) technology combined with optical remote sensing to identify hidden landslide information and analyze landslide distribution characteristics. At the same time, considering the geometric distortion of the side-looking radar image in the mountain canyon region, the paper adopts the improved R-index visibility analysis model to extract the region with good visibility for InSAR data processing, and combines the D-InSAR processing results with the optical remote sensing interpretation results to identify the hidden danger of landslide. The results show that the side-looking radar image selected in the paper can extract 48.7% of the visible area. 29 landslides are identified by consensus. Among them, 8 landslides are identified by D-InSAR technology, 28 landslides are identified by optical remote sensing technology, and 7 landslides overlap. The recognition results of the two technologies are consistent and complementary. Most of the reservoir bank landslides of Lianghekou Hydropower Station are distributed on the slope surface of 30°～45°, and the concave bank or ridge mountain with less vegetation cover is more likely to develop landslide.更多还原