Command line tool to perform landslides rapid mapping from remote sensing optical multispectral imagery using a hybrid pixel-based and object-based image analysis approach
The tool can be easily used through a docker container by building a dedicated image from the Dockerfile, that contains all the required free libraries and software dependencies.
The tool works from the command-line interface by running the following command:
Landslides_detector -o /output_folder -d /data/dem.tif -c /data/PostEvent_NDWI.tif -a /data/PreEvent_NDVI.tif -b /data/PostEvent_NDVI.tif
Additional arguments can be passed to the operator, for example:
Landslides_detector -v -t 0.2 -p "20230428_20230523" -o /output_folder -n /data/AoI.gpkg -m /data/raster_mask.tif -l /data/Soil_sealing_mask.tif -d /data/dem.tif -c /data/PostEvent_NDWI.tif -a /data/PreEvent_NDVI.tif -b /data/PostEvent_NDVI.tif -e /data/PreEvent_cloud_mask.tif -f /data/PostEvent_cloud_mask.tif
Complete operator help with description of single arguments can be displayed by running the command:
Landslides_detector -h
Input datasets required for the analysis are:
- Digital Elevation Model
- Pre-event NDVI
- Post-event NDVI
- Post-event NDWI