Appearance
OpenBuildingMapGlobal building footprint dataset
A worldwide dataset of buildings with height and occupancy types, combining OpenStreetMap, Google Open Buildings, and Microsoft ML Building Footprints.
Skip to content
Dataset
OpenBuildingMap (OBM) is a global assessment of building footprints organised in a tiled grid. For each building the dataset provides geometry, occupancy type, height, and floorspace. It covers all buildings from:
- OpenStreetMap (OSM planet dump, 1 July 2024)
- Google Open Buildings (Sirko et al., 2021)
- Microsoft Global ML Building Footprints (Microsoft, 2023)
Occupancy type and number of stories are identified using OSM land use and points of interest. Building height is estimated using the Global Human Settlement Built-up Characteristics 2023A Layer (GHSL). The dataset is described in full in the data description (PDF).
Citation
When using the data, please cite:
Oostwegel, L. J. N.; Schorlemmer, D.; Lingner, L.; Evaz Zadeh, T. (2025): OpenBuildingMap. GFZ Data Services. https://doi.org/10.5880/GFZ.LKUT.2025.002
The data are supplementary material to:
Oostwegel, L. J. N.; Schorlemmer, D.; Guéguen, P. (2025): From Footprints to Functions: A Comprehensive Global and Semantic Building Footprint Dataset. Scientific Data. https://doi.org/10.1038/s41597-025-06132-z
License: Open Data Commons Open Database License (ODbL) v1.0
File format
The dataset is distributed per country as GeoPackage (.gpkg.bz2), organised per level-6 Quadkey tile (1 271 files total). Each file contains two tables:
Building table
| Column | Type | Description |
|---|---|---|
id | Integer | Building ID. OSM ID for OSM buildings; generated for Google/Microsoft. |
floorspace | Real | Estimated floor area. |
occupancy | String | Occupancy type per GEM Building Taxonomy v2.0. |
height | String | Height / number of stories per GEM taxonomy. |
quadkey | String | Level-18 Quadkey ID (Web Mercator, EPSG:3857). |
source_id | Integer | 0 = OpenStreetMap · 1 = Google · 2 = Microsoft |
relation_id | Integer | OSM relation ID if the building is part of a relation. |
last_update | Datetime | Time of processing. |
Metadata table
| Column | Type | Description |
|---|---|---|
license | String | Always ODbL v1.0. |
number_of_buildings | Integer | Total buildings in the file. |
percentage_known_occupancy | Real | Share of buildings with a non-UNK occupancy type. |
percentage_known_height | Real | Share of buildings with a height estimate. |
percentage_known_floorspace | Real | Share of buildings with a floorspace estimate. |
percentage_source_openstreetmap | Real | Share sourced from OpenStreetMap. |
percentage_source_google | Real | Share sourced from Google Open Buildings. |
percentage_source_microsoft | Real | Share sourced from Microsoft ML Building Footprints. |
Data sources
- OpenStreetMap contributors (2024). Planet dump [1 July 2024, 00:00 UTC]. https://planet.openstreetmap.org
- Microsoft (2023). Global ML Building Footprints. GitHub
- Sirko, W. et al. (2021). Continental-Scale Building Detection from High Resolution Satellite Imagery. https://doi.org/10.48550/arxiv.2107.493.12283
- European Commission JRC (2023). GHSL Data Package 2023. https://doi.org/10.2760/098587
- Brzev, S. et al. (2013). GEM Building Taxonomy v2.0. https://doi.org/10.13117/GEM.EXP-MOD.TR2013.02
Funding
This work was funded by the European Union through the following projects:
- European Commission — Geo-INQUIRE (GA 101058518), Horizon Europe
- European Commission — The Human-Tech Nexus (GA 101073957), Horizon Europe
- European Commission — RISE: Real-time Earthquake Risk Reduction for a Resilient Europe (GA 821115), Horizon 2020
- European Commission — LEXIS: Large-scale EXecution for Industry & Society (GA 825532), Horizon 2020
- European Commission — LDT Toolbox
We like to be fully transparent about the AI tools we use. This website was built with the assistance of Claude (Anthropic).