Geelong Building Import

From OpenStreetMap Wiki
Jump to: navigation, search

Import Plan Outline

Goals

Import a full map of buildings in Geelong, from http://data.gov.au/dataset/geelong-roofprints-kml including Height and outline, plus name where published.

Schedule

  • March 2014 commencement
    • Experimental areas around Barwon Heads, Armstrong Creek
    • Ocean Grove, Point Lonsdale, Queenscliff
    • Further coastal towns and rural areas, slowly moving up towards Geelong itself

Import Data

Background

Provide links to your sources.

Data source site: http://data.gov.au/dataset/geelong-roofprints-kml
Data license: http://creativecommons.org/licenses/by/3.0/au/
Type of license (if applicable): CC-BY-A
Link to permission (if required): http://wiki.openstreetmap.org/wiki/Attribution/data.gov.au_explicit_permission
OSM attribution (if required): http://wiki.openstreetmap.org/wiki/Attribution/data.gov.au_explicit_permission
ODbL Compliance verified: yes

OSM Data Files

Import Type

Using JOSM to upload new buildings, transformed from existing shapefiles.

The source data is from the local government authority.

"Building outlines represent the roof areas as identified and measured through a combination of LiDAR analysis and photogrammetry. Additional attributes include approximate building height, in metres, and year of construction."

Data Preparation

Data Reduction & Simplification

  • Removal of 'const_yr', 'object id' and other unneeded keys
  • Removal of ways with height <= 0.0

Alignment

Mostly the data appears aligned to GPS traces, though is offset from Bing.

Example: http://www.openstreetmap.org/changeset/19821285

You can see the offset from the GPS trace to the Surf Life Saving Club is approximately aligned to the buildings.

Martin von Wyss, dataset curator says:

"I spent many hours testing the data in all of the formats and I'm now certain the problem isn't with the data or the transformations that happened on data.gov.au. The problem seems to be that orthophotos in Bing Maps are misaligned at varying zoom levels."

Tagging Plans

  • building=yes
  • height=12.34
  • name=Example Building

Changeset Tags

JOSM will handle the majority of this.

Changeset messages will be 'Building import from data.gov.au for Geelong'

Data Transformation

Describe the transformations you'll need to conduct, the tools you're using, and any specific configurations or code that will be used in the transformation.

Data Transformation Results

Examples:

Full data:

Data Merge Workflow

Team Approach

Team:

  • User:CloCkWeRX (Import, misc armchair mapping)
  • User:geesona (Geelong based mapper, verification of data, misc GPS data)
  • Interested members of the talk-au mailing list.

References

List all factors that will be evaluated in the import.

Workflow

  • Download original shapefiles (done)
  • Transform to .osm via JOSM (done)
  • Manipulate xml with text editor (done)
  • Reload into JOSM
  • Add mapnik as a background layer
  • Import only new buildings by hand.

Conflation

Existing buildings will be either kept, merged, or discarded depending on the quality of mapping. The more metadata or if surveyed, the more likely it will be preferred over the lidar version.

A separate effort to add building height by hand to existing buildings may take place at a later stage.

QA

  • Keepright and JOSM validations of the dataset for overlapping buildings.
  • Verification of data from local knowledge