Landslide extraction utilizing optical satellite images with deep learning

Abstract

The 2018 Hokkaido Eastern Iburi Earthquake brought about great numbers of landslides in the broad area of the eastern Iburi region, Hokkaido, especially Atsuma town. As emergency disaster response to the earthquake, PASCO Corporation carried out 1) taking oblique aerial photographs, 2) automatic extraction of landslides using optical satellite imagery (SPOT6), 3) detection of surface deformation with DInSAR analysis and 4) field surveys based on DInSAR analysis. This paper introduces the analysis of extraction utilizing optical satellite images with deep learning, which has been actively studied in recent years.