AI building-level damage classification from pre/post-disaster RGB imagery, using a Siamese U-Net ensemble running on ONNX Runtime.
beaconGIS turns pre- and post-disaster satellite imagery into a usable damage map in minutes. Load your two scenes, draw an area of interest, click Detect — get a colour-coded building footprint layer classifying each building as No Damage, Minor, Major, or Destroyed.
Built for disaster-response teams, humanitarian mapping, insurance damage assessment, and reconstruction planning. Trained on 19 real disaster events across 6 disaster types (hurricanes, earthquakes, wildfires, floods, tsunamis, volcanic eruptions).Specially fine tuned for generalization across different building types and disaster events.
You can save the classification layer as a geopackage.
You can save the binary mask of the classification for training your own models.
Plugin auto-generates a draft SitRep report you can edit and share.
You can grab post disaster images quickly from the satellite imagery downloader.
Plugin Tags