IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium | 2019
Mountainous Landslide Recognition Based on Gaofen-3 Polarimetric SAR Imagery
Abstract
Full-polarimetric SAR image is very useful for the landslide monitoring especially in the well vegetation-covered mountainous region. GaoFen-3 (GF-3) is the first civil C-band fully polarimetric SAR satellite in China. In order to test the ability of GF-3 full-polarimetric SAR data on landslide monitoring, 2 cases of mountainous landslide in southwestern part of China in 2017 and 4 polarimetric decomposition methods (Pauli, Krogager, Freeman, and H-α/A) were selected. It was found that GF-3 full-polarimetric SAR data had a good performance on that. The scattering mode had changed from volume to surface mode by the four methods which indicated the vegetation covered has been destroyed by landslide. In the incoherent decomposition methods, the scattering mode in the vegetation region around the landslides had a more consistent result than that in the incoherent decomposition methods. It can also be a good tool to do the autonomous landslide recognition in a larger region combined with some change detection methods.