Cognised existence: A building is a roofed, permanent physical structure that shelters human activity. It has use, age, size, and geometric footprint, and these properties may come from different sources.

Question: What buildings exist at a location and what are their characteristics?

OSM wiki: https://wiki.openstreetmap.org/wiki/Key:building


Realisations

OpenStreetMap — building=*

OSM contains building footprints globally. Coverage is excellent in urban areas and variable in rural and developing regions.

Primary tag: building=yes (generic), or typed values such as building=residential, building=commercial, building=school, building=hospital.

osmnx access

import osmnx as ox
ox.settings.cache_folder = ".cache/"
 
buildings = ox.features_from_place(
    "Copenhagen, Denmark",
    tags={"building": True}
)
# Returns GeoDataFrame, geometry = Polygon/MultiPolygon, CRS = EPSG:4326

Geofabrik layer

gis_osm_buildings_a_free_1.shp — polygon layer, fields: osm_id, code, fclass, name.

Overpass (fallback)

[out:json][timeout:90];
area["name"="Copenhagen"]["admin_level"="7"]->.a;
(way["building"](area.a); relation["building"](area.a););
out body; >; out skel qt;

Key OSM attributes

TagMeaning
building=*Building type (residential, commercial, industrial, school)
building:levels=*Number of above-ground floors
start_date=*Construction year (sparse)
name=*Building name (public buildings)
addr:housenumber, addr:streetAddress tags (variable coverage)

Geometry Representations

Rep IDSource DatasetGeometry TypeNative CRSSuitable ForNot Suitable For
buildings_osm_footprintOpenStreetMapPolygonEPSG:4326Footprint mapping, area calculation, building count, urban morphologyRegister-quality attributes (year, use code, heating) are sparse in OSM
buildings_osm_centroidOpenStreetMap (derived)PointEPSG:4326Density mapping, point-in-polygon, proximity countsArea statistics, shadow and solar analysis

Note: OSM building footprints are the best globally available open polygon source. Attribute quality varies strongly by region and editing community.


Limitations

  • Attribute completeness varies widely by country and urban density.
  • Year of construction (start_date) is sparse globally.
  • Building type tags are inconsistent across regions.