Wisconsin Multi-County Building Footprint Import

From OpenStreetMap Wiki
Jump to: navigation, search


Import building footprints from the State Cartographer's Office into OSM.


Import is in planning and data transformation stages. Actual imports will occur beginning in December 2018 and proceeding through the first half of 2019.

Import Data

All data come from the Wisconsin State Cartographer's Office.


SCO GeoData: http://geodata.wisc.edu/opengeoportal/
Type of license: Public Domain
OSM attribution: Wisconsin State Cartographer's Office
ODbL Compliance verified: yes


The following is a list of all counties currently part of the import. Counties not listed were either unavailable, or did not have data of sufficient quality to import.

  • Ashland
  • Bayfield
  • Brown
  • Buffalo
  • Calumet
  • Clark
  • Dane
  • Eau Claire
  • Iron
  • Kenosha
  • Lincoln
  • Menominee
  • Milwaukee
  • Oconto
  • Oneida
  • Outagamie
  • Ozaukee
  • Sauk
  • Vilas
  • Winnebago
  • Wood

OSM Data Files

Most OSM files have been added to a public Google Drive folder here, and the rest will be added as they are completed.

Import Type

This is planned to be a one-time import, which will be conducted via JOSM.

Data Preparation

Data Reduction & Simplification

Schema is highly variable between counties. When possible, attributes will be mapped to OSM tags, but will vary county to county. Preliminary review of the shapes shows that footprints are already at an ideal level of detail for importing.

Tagging Plans

Unless noted in this section, county footprints contain no useful attributes.

Specific Counties


Several thousand features contain a "BldgHeight" value in feet. These will be converted to meters and included as height=* tags.


The fields "BuildingUse" and "BuildingClass" can be used to identify more specific building=* tags. "BuildingUse" includes values like 'Industrial', 'Residential', and the like. "BuildingClass" falls into either 'Primary' or 'Accessory', helpful in determining structures which might fall under 'farm_auxilliary', for instance.


Sauk County footprints have "Building Height", which will need to be converted to meters. A handful of features also include "Site Address" information, which can be split into both addr:housenumber and addr:street tags.


Nearly all Vilas County footprints include an "Address" field. Similar to Sauk County, these can be split into add:housenumber and addr:street tags.


Winnebago footprints have a "Building Use" field. Many of the values in this field have direct equivalents in OSM building tags, such as 'residential', 'garage', 'silo', etc.

Changeset Tags

Source=Wisconsin State Cartographer's Office

Data Transformation

  1. Convert to OSM, using ogr2osm, JSON, etc.
  2. Remove extraneous tags, add building=yes.
  3. Download all buildings for area being worked on (county, tract, etc).
  4. Identify overlapping/building-in-building validation errors, then remove any version:0 ways.
  5. Remaining buildings should be checked against full OSM dataset for additional validation areas prior to uploading.

Data Transformation Results

In progress.

Data Merge Workflow

Team Approach

This project will be managed through US Tasking Manager. https://tasks.openstreetmap.us/project/108

County Progress
County Username Status
Milwaukee shuii Complete
Kenosha jdcarls2 Complete
Ashland jdcarls2 Complete
Bayfield jdcarls2 Complete
Dane JohnHoverCraft In Progress
Oneida JohnHoverCraft In Progress
Sauk jdcarls2 In Progress
Outagamie OttoShade In Progress
Winnebago OttoShade In Progress
Calumet OttoShade In Progress


  1. Load data file.
  2. Perform initial validation to ensure there are no errors in import data.
  3. Download OSM data from same area.
  4. Perform second validation.
  5. If applicable, select all ways which were identified as "crossing buildings".
  6. Filter selection by searching for "version:0". These will be import buildings which intersect with OSM buildings.
  7. Delete selection.
  8. Perform final validation. Ensure that there are no building-related validation errors that involve import data.
  9. Upload.

Data will be broken up into city blocks or township sections as needed to keep changeset sizes within several thousand objects.

A detailed guide to the workflow can be found in the shared Google Drive folder.


Objects intersecting with existing OSM buildings are removed from the import data.


Following successful import of the data, a random subset will be evaluated against imagery and existing OSM data