Combining Multi-Spectral Satellite Imagery and Machine Learning Methods to Establish Automatic Identification System of Landslides and Factor Analysis of Typhoon Events( I )

Project Details


Due to its special geographical location, its geology, topography and climate, Taiwan's typhoon and plum rains often cause landslides and mudflows. In recent years, the frequency of extreme rainfall events has gradually increased, and the types of disasters caused by extreme rainfall have changed from only floods in the past to large-scale floods and landslide compound and complex disasters. For example, Typhoon Morakot in 2009 and Typhoon Soudelor in 2015. With the advancement of spatial information technology, satellite remote sensing images are often used as slope disaster monitoring because of their wide range of sensing and short sensing periods. In particular, in recent years, FORMOSAT-2 and FORMOSAT-5 made in Taiwan have provided satellite imagery of Taiwan in higher spatial and temporal resolution. Therefore, this study attempts to establish a typhoon event landslide automatic recognition system based on multi-spectral satellite imagery and various machine learning methods, and cooperate with the National Space Organization (NSPO) of the National Applied Research Laboratories to apply and promote FORMOSAT-2 and FORMOSAT-5 data, and develop the method of rapid identification before and after the occurrence of landslide disasters. At the same time, after the establishment of the landslide identification system, the historical landslide type is analyzed through satellite imagery. Secondly, through the Typhoon Type Index developed by our lab, typhoon characteristics and Rainfall pattern classification, the effects of different types or conditions on the spatial distribution of landslides were discussed. Furthermore, the prediction model of the spatial and scale distribution of landslide disasters caused by typhoon characteristics and rainfall characteristics is further established. It is expected to be used for landslide disaster identification, disaster information analysis, disaster prevention and response, and disaster risk warning application.
Effective start/end date1/11/1931/10/20

UN Sustainable Development Goals

In 2015, UN member states agreed to 17 global Sustainable Development Goals (SDGs) to end poverty, protect the planet and ensure prosperity for all. This project contributes towards the following SDG(s):

  • SDG 11 - Sustainable Cities and Communities
  • SDG 15 - Life on Land
  • SDG 17 - Partnerships for the Goals


  • Typhoon events
  • landslide factors
  • satellite image recognition
  • machine learning


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.