Across Berlin we mapped the urban surface into nine primary classes, enriched with status attributes — a semantically rich picture of the city's transport corridors and the land covers around them.
The challenge
The client needed a high-resolution map that could distinguish, in one pass, between road, shoulder, parking, bikeway, footway, access ways and keep-out areas — and still handle the messy reality of construction zones, occlusions and ambiguous edges that defeat conventional methods.
Our approach
We paired high-resolution aerial and satellite imagery with a rule-aware AI inference engine, classifying each region into nine primary polygon classes:
- road, road shoulder, parking area, access way
- bikeway, footway, railroad bed, keep-out area, water
On top of the geometry, each feature can carry quality-aware flags — construction, traffic island, elevated, ambiguous/difficult, or fully occluded (“invisible”) — so downstream analysis knows exactly how much to trust each polygon.
Results & benefits
- A unified, nine-class polygon map of Berlin's transport network and water bodies.
- Clear separation of road, shoulder, parking, access ways, bikeway, footway, railroad and keep-out areas.
- Status flags that make the data quality-aware rather than blindly trusted.
- A fully automated, scalable workflow suited to city-wide annual updates.
