Skip to content

Latest commit

 

History

History
53 lines (41 loc) · 2.62 KB

README.md

File metadata and controls

53 lines (41 loc) · 2.62 KB

Bellingham-Bldg-Addr

Not ready for import

Building and Address Import

This project will import data from the City of Bellingham, WA. Bellingham provides a wide array of data from their website, including docks, sidewalks, trees, fire hydrants, roads, and of course addresses and building outlines. Please feel free to offer suggestions and PR to improve the process.

The import process will use the US Tasking Manager. Bellingham precincts will be used to break the import into 67 small tasks.

License

The City of Bellingham offer open data with no strings attached.

Process

This process is built on using PostGIS and Python. It has been developed using Fedora 24. It should work on MacOS. To use the process, clone the repository and modify the import.sh and bellingham_osm.sh scripts to define your working directory and access to Postgresql. Before running the scripts, add the functions below to your PostGIS/Postgresql server. They are needed to convert the road names into ones that OSM like.

  • run import.sh to build the Postgresql tables
  • run bellingham_osm.sh to build individual .osm files for import

Requirements

  • PostGIS
  • Python
  • Postgresql functions from my sql repository
expand_directions.sql
expand_roads.sql
format_roads.sql
fullname.sql

Whatcom Roads

Whatcom roads are available as a background layer. The WMS service can be loaded in JOSM using the images preferences with the following url

http://tiles.snowandsnow.us/geoserver/wms?service=WMS&request=GetMap&styles=&TRANSPARENT=TRUE&bbox=-1.37021663844817E7,6214544.26722411,-1.35320129795646E7,6275268.80055685&width=768&height=330
  1. Press Get Layers, select Washington Roads and
  2. Change the image type to png (This creates a transparent background so aerial imagery is visible)
  3. Name the layer Whatcom Roads

Notes

This is my first attempt using parcel data to help identify address nodes to building outlines. One of the problems that needed to be overcome was buildings that covered more than one parcel. In some cases it appeared that either the building outline was shifted or that the parcel outline was incorrect.

Just Over

Below is an example of a building in the middle of two parcels

Middle

Query to select the correct parcel. The percentage, I used 80% seem right for this use.

UPDATE bellingham_bldg SET parcel_cod = b.gid
FROM bellingham_parcels p join bellingham_bldg b ON ST_INTERSECTS(b.geom,p.geom)
WHERE b.parcel_cod IS NULL AND ST_AREA(ST_INTERSECTION(p.geom, b.geom)::geometry)/ST_AREA(b.geom::geometry) > .8;