High-Level Project Summary
Natural disasters have existed forever, but only in recent decades, due to population growth, these are having a greater impact on the quality of human life. This project aims to design a processing model of data, combining GIS tools with the power of artificial intelligence systems, in order to be able to predict possible future cases of landslide disasters, and thus alert the population most at risk of be affected to avoid suffering the consequences.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
Using The Cooperative Open Online Landslide Repository (COOLR) reports, the landslide data was filtered according to site and date accuracy.
With this smaller database, satellite images (Sentinel-2) of the places were obtained on dates prior to the disaster.
These images were processed with Google Earth Engine to extract information that can facilitate the prediction of future cases. Data such as the Normalized Difference Vegetation Index, Normalized Differential Water Index and False color for detection of urban areas were obtained.
Then this data was exported to be processed by a neural network, aiming to be able to predict future cases.
A set of the data used to test the model is available for your review. Due to serious equipment drawbacks, not all landslide data was processed; but the filtered catalog and the code used in Earth Engine are left available.
TOOLS USED FOR THE PROJECT
- Google Earth Engine
- Visual Studio Code
- COOLR Landslide Viewer
USED PROGRAMMING LANGUAGES
- Python
- JavaScript
Space Agency Data
- NASA landslides resources - For the dataset of Landslides
- IMERG: Integrated Multi-satellitE Retrievals for GPM | NASA Global Precipitation Measurement Mission - Through Google Earth Engine images of the GPM, landsat 8 and sentinel 2 were used for data extraction.
Hackathon Journey
...
References
- Sentinel-2 MSI: MultiSpectral Instrument, Level-1C
- Cooperative Open Online Landslide Repository (COOLR)
- Landslide Prediction with Model Switching
Darmawan Utomo , Shi-Feng Chen and Pao-Ann Hsiung
Computer Science and Information Engineering, National Chung Cheng University
TOOLS
- Google Earth Engine
- VS CODE
Global Judging
This project has been submitted for consideration during the Judging process.