Flood mapping: Datasets & Models

Manually digitizing flood boundaries from imagery is slow, subjective, and hard to scale when a disaster spans hundreds of square kilometers and imagery arrives faster than analysts can process it. Automating segmentation lets geospatial teams generate consistent flood extent maps within minutes of a drone flight or satellite pass, feeding directly into GIS layers for emergency response, insurance claims, and infrastructure triage. For drones and geospatial operators, this means faster situational awareness and repeatable analysis across repeated survey missions.

Instance segmentation delivers pixel-accurate water masks that support quantitative area calculations, not just detection flags. The workflow integrates with existing drone fleets, orthomosaic pipelines, and satellite data sources, and scales from a single site survey to region-wide monitoring. Once trained on a representative flood dataset, the model runs on new imagery automatically, cutting turnaround time on flood mapping deliverables.

1. Upload an image: Drop any photo into the widget above to try flood mapping for free, right on this page.
2. Preview instant results: The model detects flood water, building, road, vehicle, vegetation within seconds and returns an annotated image with structured predictions, so you can immediately see how well it works. Adjust the detection classes and re-run until the results fit your scenario.
3. Like what you see?: With a free Roboflow account, you can deploy this workflow with the Serverless Cloud API, customize it in your workspace, or train RF-DETR (instance segmentation) on your own images for production accuracy.
4. Get started for free: No code or training data required to begin: upload, review the results, and make it part of your stack in a single click.