Semi-supervised cross-domain feature fusion classification network for coastal wetland classification with hyperspectral and LiDAR data

• DOI: 10.1016/j.jag.2023.103354
• Publication Date: 2023-05-19T17:11:11Z
• Authors: Fangming Guo; Zhongwei Li; Qiao Meng; Guangbo Ren; Leiquan Wang; Jianbu Wang; Huawei Qin; Jie Zhang
• Keywords: Feature (linguistics); Sensor Fusion

Multi-source remote sensing monitoring plays a crucial part in the ecological protection and restoration of coastal wetlands. However, due to inaccessible wetlands environment, lacking labeled samples is challenge wetland classification. In this article, an unsupervised cross-domain feature fusion supervised classification network (UF2SCN) proposed for classification, which fuses hyperspectral image (HSI) light detection ranging (LiDAR) data. First, single branch end developed get HSI LiDAR feature, extraction model with spectral attention deployed obtain average distribution characteristics all samples, data utilized guide whole process. Second, spatial applied used uses limited samples. Finally, two stages training strategy improve ability fusion. Experiments conducted on datasets created by ourselves prove validity method wetland.