摘要
本文采用GF-6 WFV影像和Landsat-8-OLI影像,结合单波段阈值法、植被指数法和决策树算法,通过计算多种指数并确定阈值,构建决策树提取规则,对小浪底水库进行提取。提取结果表明:基于GF-6影像的NDWI和MSWI方法提取结果最佳,总体精度为0.9967、F1_score为0.9968、kappa系数为0.9934。说明国产GF-6卫星遥感影像数据在小浪底水库提取中具有较高应用价值。
This article uses GF-6 WFV images and Landsat-8-OLI images,combined with single band threshold method,vegetation index method,and decision tree algorithm,to calculate multiple indices and determine thresholds,construct decision tree extraction rules,and extract the Xiaolangdi Reservoir.The extraction results show that the NDWI and MSWI methods based on GF-6 images have the best extraction results,with an overall accuracy of 0.9967,the F1_score is 0.9968,and the kappa coefficient is 0.9934.The domestic GF-6 satellite remote sensing image data has high application value in the extraction of Xiaolangdi Reservoir.
作者
吴宗泽
李嘉
WU Zongze;LI Jia(Map Institute of Guangdong Province,Guangzhou 510030,China)
出处
《测绘与空间地理信息》
2026年第2期163-165,共3页
Geomatics & Spatial Information Technology
