Knowledge of building density is vital for policy-making, community planning, understanding environmental impacts, delivering services and responding to emergencies, but current information is often unavailable in countries that are low on resources.
Building footprints are a key ingredient for estimating population density, so Google teams trained a deep learning model to determine the footprints of buildings from high resolution satellite imagery. The AI model detects buildings in dense areas and the most rural areas, which can be challenging due to unique shapes and materials. Each building detected also includes a Plus Code that acts as an address, provides location information, and estimates the size of the building.
The data set contains 817 million building detections across Africa, South Asia and South-East Asia, allowing for better community planning and emergency response. Open Buildings supports governments, nonprofits, researchers and the public with applications such as disaster response, household survey planning, electricity provision and urban planning.
“When I lived in Rwanda for two years, there was unprecedented rainfall that destroyed fields and homes. I imagine this data set could help governments and orgs better identify high-risk areas and offer potentially life-saving assistance.”
- Lily Adelstein, Project Manager at Telepath
The site could not be loaded because of an error, please check your internet connection and/or reload the page.
Your browser is outdated and may not display all features of this site.