Scope:
100,000 images, 2 million polygons.
Projects:
Flood Detection | Building Count | Urban Developememt
Challenge:
Scope and complexity of the task, which comprised identifying buildings’ borders in dense urban areas, in a low visibility context.
Solution:
Dividing AOIs in smaller tiles to allow for better zooming, using several basemap layers for comparison and using desktop quality control processes allowed us to deliver a high sensitivity dataset.
Use of Building Footprint
Flod Detection:

Detecting key objects (water, road, building) disaster planning and evacuation plans.
Segmentation of debris for airplane strip check.
Helmets and gear in a factory.

Building Count:

Detecting key objects (water, road, building) disaster planning and evacuation plans.
Segmentation of debris for airplane strip check.
Helmets and gear in a factory.

Urban Developememt:

Detecting key objects (water, road, building) disaster planning and evacuation plans.
Segmentation of debris for airplane strip check.
Helmets and gear in a factory.