Detection of Landslides from High Resolution Optical Satellite Data

Landslide is a major natural disaster in India or countries elsewhere where significant portion of landmass is in mountainous areas. For such areas, identification of landslide forms a fundamental input to develop landslide hazard management and mitigation measures. Previously, landslides used to be identified either by manual image interpretation process or by field investigations that demanded considerable amount of time and labour. Recently, a new technique using advanced object-based image classification method is developed to automatically detect landslides from satellite data. This technique imitates the human interpretation process in identification of landslides. This technique has the potential to be integrated in the DMS programme of ISRO for landslide incidence and hazard mapping.