Microsoft Building Footprint Data - OpenStreetMap Wiki (2024)

Microsoft has released multiple building footprint datasets, all of which are licensed under the same Open Database License (ODbL) as OpenStreetMap. Any import of these building footprints must strictly follow the import guidelines.

Microsoft Building footprints are available in most of the world^[1], including USA,^[2] Canada,^[3] Tanzania&Uganda,^[4] and Australia^[5].

The building footprints can be added manually to OpenStreetMap using the RapID editor or the mapwithai plugin for JOSM.

In March 2017, Microsoft released an initial dataset containing approximately 9.8 million high-quality building footprints with heights in 44 U.S. states. Key selected metropolitan areas are covered. The footprints are the result of Microsoft's efforts and not purchased or obtained from other sources. They are freely available for download for import into OpenStreetMap. Even if your locality has buildings, it worth checking out these high quality footprints even if just to capture the building heights.

In October 2020, Microsoft released a dataset containing 11.3million buildings in Australia.

In 2022 Microsoft released their Global ML footprints. A visual heat-map of the area this database covers is shown in their diagram.

The height data is measured in meters, the default unit for OSM height measurements. No unit conversion is needed.

Approx. 9.8 million building footprints for portions of metro areas in 44 US States in Shapefile format.

Hand digitized footprints from very high resolution aerial photography captured by Microsoft. The Height attribute was interpolated from a digital terrain model derived from the same data.

The footprints were digitized in 2015 from imagery captured in 2014 & 2015.

def filterTags(attrs): if not attrs: return tags = {} tags['building'] = 'yes' if 'HEIGHT' in attrs and attrs['HEIGHT'] != '': tags['height'] = attrs['HEIGHT'].rstrip('0') return tags

California has more than triple the amount of data available than any other state. Importing it will be no small task but doing it in chunks by several people will make it manageable. The buildings in the Bay Area alone in the file stretch from Clear Lake way down to Hollister and run along the coast in Santa Cruz all the way up the East Bay. Several other large cities and areas are included in the data. There have been several imports of buildings here in California and many people have put in a lot of work tracing individual buildings. These data will tie in those imports and will be a valuable addition. Here are links to the data broken apart further by region along with the status of buildings in the areas.

There's several scripts for outputting some information, generating modified osm data and extracting buildings that aren't present at all in OSM..

Only tested against the Michigan data which has 8140 geometries, concentrated in the Detroit area. Not sure how things will go with larger datasets. A clipped out region of reasonable size should work fine (the scripts complete in a few seconds for ~10,000 buildings on my older laptop).

For Michigan/Detroit, the main takeaway is that a *lot* of the buildings are already in OpenStreetMap.

This histogram is calculated using the largest single overlap for each existing OpenStreetMap building (data at

The areas there are presently calculated stupidly, in WGS 84, but I think the relative information should be fine. So of the ~14422 buildings present in OSM, 7630 don't overlap the Microsoft buildings at all and 5 or 6 thousand overlap quite a lot. Of the visual review I've done, I'd say that the existing OSM buildings tend to be more detailed and line up more closely with current Bing Imagery.

Separately, there are 410 buildings in the Microsoft data that do not exist in OSM (take a look

A more sophisticated matching algorithm is probably a good idea, but for the Detroit data I would be pretty comfortable mechanically adding

the heights for the buildings where the overlap is roughly 80% or higher and then doing a more manual process for the new buildings (checking against newer imagery?), and then also doing some sort of more manual process to capture the information from the several hundred buildings with smaller overlaps.

This dataset was already discussed on imports mailing list. Some areas seems to have better quality than others, and round buildings are transformed to squares.

Approx. 125 million building footprint polygon geometries in all 50 US States in GeoJSON format.

Computer generated.

Data Vintage:
Bing Imagery is a composite of multiple sources, and it is difficult to know the exact dates for individual pieces of data.

def filterTags(attrs): if not attrs: return tags = {} tags['building'] = 'yes' return tags

In October 2020, Microsoft released 11,334,866‬ buildings in Australia extracted from 2013–2018 Maxar imagery and made them available for editing in RapiD.

In 2022 Microsoft released 856 million polygons of building footprints. This is available through RapiD. It covers many areas of the globe. It is available at GitHub.