This project was mainly aimed at demonstrating "AI-Mapping" leveraging AI technologies to analyse and digitize aerial and satellite imageru. As soon as there are more tools and more open data available we reviewed the concept to share specific datasets with the community. It is dedicated to analysis of urban environment and land resources as well as to the wide range of educational and resesrch projects.
Region | Name | Area | Download |
---|---|---|---|
Central Asia | 5 cities - central areas | 5*100km² | Geojson, 84Mb GeoPackage, 66Mb |
👉 Download this PDF showing green patterns in European cities
Region | Name | Feature Count | Format | Size (zipped) | Last update |
---|---|---|---|---|---|
Uzbekistan | Admin boundaries with population | ~138 | GeoPackage | 1.3MB | |
Uzbekistan | Kontur population 500m | ~108,000 | GeoPackage | 6.5MB | |
Tashkent | Geoalert building footprints | ~387,000 | GeoPackage | 57MB | |
Tashkent | Geoalert construction detections | ~515 | GeoJSON | 114KB | ~2023 |
Uzbekistan | Geoalert building footprints, 4 cities | ~301,000 | GeoPackage | 31MB | |
Uzbekistan | MS building footprints | ~6,500,000 | GeoPackage | 684MB | |
Uzbekistan | Places with population, Openstreetmap | ~4,800 | GeoPackage | 2MB |
If you are an OpenStreetMap contributor, you can use this data to contribute to OSM directly or accelerate your own mapping efforts. Since the data was generated automatically, it should be thoroughly validated before import. You can preview and edit the data using QGIS, the OSM ID editor ("custom Map Data"), or JOSM. We also strongly recommend that you check OSM imports community guidelines. We've created the Open Urban Mapping wiki - you can find some tips on editing and doing imports based on our datasets.
The auto-mapping approach can also be used to help create maps from scratch in times of emergencies such as in Irkutsk region, Russia, that was heavily flooded in the summer 2019, so we appreciate any contribution and/or related data requests.
If you'd like to help us with documentation, integration of datasets into third-party applications like JOSM, RapID, etc., or promote this project, please check out the issues or create one to submit your request. You can also contact us directly at [email protected]
Open datasets are created based on "Mapbox Satellite" in order to be compatible with the OpenStreetMap license and contribution guides.
All data in this project is licensed under the Open Database License (ODbL). The input imagery data is copyrighted by the data providers but is not distributed along with the dataset.
You are free to copy, distribute, transmit and adapt this data, as long as you credit "Geoalert / Mapbox" as the data source. You are supposed to distribute the derived data under the same license.
- Coordinate reference system - EPSG: 4326
- Data format - GeoPackage or GeoJSON
- Moscow and Moscow region update, Jul 23 2021: 1M+ building footprints (in Russian)
- Chechnya update, Jul 16 2021 - the least mapped region in OSM
- Subscribe to Geoalert blog
- Mapflow.AI - AI mapping and imagery analysis platform
- Microsoft building footprints
- RapID - Facebook editor for OpenStreetMap
Date | Description |
---|---|
2024-07-18 | Remove Russia datasets as out-of-dated. Update Tashkent construction areas, data relevance – Sep, Oct, 2023 |