基于三种高分辨率影像样本迁移学习的城市水体提取研究

史佳睿,申茜,彭红春,李利伟,姚月,汪明秀,王茹

Research on Urban Water Body Extraction based on Transfer Learning of Three High-resolution Image Datasets

Jiarui Shi,Qian Shen,Hongchun Peng,Liwei Li,Yue Yao,Mingxiu Wang,Ru Wang

表1 遥感影像参数

Table 1 Remote sensing image parameters

影像名称 波段 融合后空间分辨率/m 成像时间 中心点经纬度 像元数量 Worldview-2 R,G,B,NIR 0.5 2017/9/19 116.7N,39.8E 16 022×32 769 Superview-1 R,G,B,NIR 0.5 2017/8/22 116.8N,39.7E 11 318×11 734 Geoeye-1 R,G,B,NIR 0.5 2017/5/20 116.7N,39.8E 11 214×11 637 BJ-2 R,G,B,NIR 0.8 2019/6/24 116.8N,39.9E 36 610×36 518 GF-2 R,G,B,NIR 0.8 2018/9/05 116.7N,39.9E 34 431×33 498 GF-1 R,G,B,NIR 2 2019/7/12 116.7N,39.9E 21 770×21 291 GF-1B R,G,B,NIR 2 2019/9/24 117.0N,39.7E 40 269×40 028 GF-1C R,G,B,NIR 2 2020/3/11 117.0N,39.7E 39 908×39 725 GF-1D R,G,B,NIR 2 2019/10/21 116.7N,39.7E 41 347×40 289