Microsoft Building Footprints

✦ DISCOVERED BY ORIONAICOLLECTIONMicrosoftBuildings and infrastructureGlobal
Map preview of Microsoft Building FootprintsGlobal

About this dataset

Bing Maps is releasing open building footprints around the world. We have detected over 999 million buildings from Bing Maps imagery between 2014 and 2021 including Maxar and Airbus imagery. The data is freely available for download and use under ODbL. This dataset complements our other releases. For more information, see the GlobalMLBuildingFootprints repository on GitHub. ## Building footprint creation The building extraction is done in two stages: 1. Semantic Segmentation – Recognizing building pixels on an aerial image using deep neural networks (DNNs) 2. Polygonization – Converting building pixel detections into polygons **Stage 1: Semantic Segmentation** !Semantic segmentation **Stage 2: Polygonization** !Polygonization ## Data assets The building footprints are provided as a set of geoparquet datasets in [Delta][delta] table format. The data are partitioned by 1. Region 2. quadkey at [Bing Map Tiles][tiles] level 9 Each `(Region, quadkey)` pair will have one or more geoparquet files, depending on the density of the of the buildings in that area. Note that older items in this dataset are *not* spatially partitioned. We recommend using data with a processing date of 2023-04-25 or newer. This processing date is part of the URL for each parquet file and is captured in the STAC metadata for each item (see below). ## Delta Format The collection-level asset under the `delta` key gives you the fsspec-style URL to the Delta table. This can be used to efficiently query for matching partitions by `Region` and `quadkey`. See the notebook for an example using Python. ## STAC metadata This STAC collection has one STAC item per region. The `msbuildings:region` property can be used to filter items to a specific region, and the `msbuildings:quadkey` property can be used to filter items to a specific quadkey (though you can also search by the `geometry`). Note that older STAC items are not spatially partitioned. We recommend filtering on items with an `msbuildings:processing-date` of `2023-04-25` or newer. See the collection summary for `msbuildings:processing-date` for a list of valid values. [delta]: https://delta.io/ [tiles]: https://learn.microsoft.com/en-us/bingmaps/articles/bing-maps-tile-system

Imagery details

Related datasets

✦ Suggested by OrionAI
Building footprints: Asia (Microsoft)

AI-derived building footprint polygons for Asia from Microsoft's Global ML Building Footprints, in 749 quadkey tiles of line-delimited GeoJSON, uploaded 2026-02-23.

Buildings and infrastructure · Microsoft
Building footprints: Bhutan (Microsoft)

AI-derived building footprint polygons for Bhutan from Microsoft's Global ML Building Footprints, in 17 quadkey tiles of line-delimited GeoJSON, uploaded 2026-02-23.

Buildings and infrastructure · Microsoft
Microsoft Building Footprints for Kentucky - Features

From: MS Buildings Microsoft recently released a free set of deep learning generated building footprints covering the United States of America. In support of this great work and to make these building footprints available to the ArcGIS community, Esri has consolidated the buildings into a single layer and shared them in ArcGIS Online. The footprints can be used for visualization using vector tile format or as hosted feature layer to do analysis.  Learn more about the Microsoft Project at the Announcement Blog or the raw data is available at Github.

Buildings and infrastructure · Kentucky_DGI
Building footprints: Belize (Microsoft)

AI-derived building footprint polygons for Belize from Microsoft's Global ML Building Footprints, in 9 quadkey tiles of line-delimited GeoJSON, uploaded 2026-02-23.

Buildings and infrastructure · Microsoft
Building footprints: Bahrain (Microsoft)

AI-derived building footprint polygons for Bahrain from Microsoft's Global ML Building Footprints, in 3 quadkey tiles of line-delimited GeoJSON, uploaded 2026-02-23.

Buildings and infrastructure · Microsoft
Building footprints: Bangladesh (Microsoft)

AI-derived building footprint polygons for Bangladesh from Microsoft's Global ML Building Footprints, in 53 quadkey tiles of line-delimited GeoJSON, uploaded 2026-02-23.

Buildings and infrastructure · Microsoft

Indexed by OrionGIS from Microsoft Planetary Computer. Data is loaded from the publisher's own server when you use it. Permanent link