OpenStreetMap road network parser

Generate equidistant points along urban area road network

This repo. contains code + a set of scripts for extracting equidistant points
for cities/urban areas in a national road network. Specifically, for each road
within a city, a set of points are generated at equal distances along with the
associated heading/bearing for each point. The points/bearings are then output
to a CSV.

This is part of a larger project which uses these sample points and associated
bearings to query
Google street view API
for all images on the left and right hand side of a road network. The sample
points may be useful as inputs into other projects or may be used in isolation
to derive city specific geospatial features. For example, we have
found that the distribution of sample
point bearings for most UK cities is biased toward North, East, South, West and
not uniform:

This could be related to historical relationships with neighboring towns and
is representative of the overall spatial layout of a town’s road network. We
plan to explore this idea/finding at a later date. For example, it may be
interesting to compare this data with road safety: towns with a predominant
East, West road orientation may exhibit higher occurrence of incidents with
respect to traveling in the direction of a rising/setting sun.

Dependencies

• osmium
• nodejs, npm
• osmtogeojson
• geojsplit
• json_pp

osmium

mkdir osmium ;cd osmium
git clone https://github.com/mapbox/protozero
git clone https://github.com/osmcode/libosmium
git clone https://github.com/osmcode/osmium-tool
cd osmium-tool
mkdir build ;cd build
cmake ..
make
sudo make install

geojsplit + osmtogeojson

sudo npm -g install geojsplit
sudo npm -g install osmtogeojson

Building

make

Running

The whole process can be executed by invoking the run.sh script.

./run.sh

This will perform the following steps:

1) Get ONS major towns and cities polygon data.
data/area_geojson will download the 112 major
towns and cities geojson data from
ONS Geography portal,
then extract each city into individual geojson files in data/area_geojson/split/.

2) Extract the road network for each city.
extract_roads.sh will download the latest open street map
data for Great Britain, extract the road network using
osmium, and then for each city,
extract the road network bounded by the city polygon obtained in step (1) and
convert to geojson:

3) Generate sample points for each city.
For each individual city road network in data/city_roads_geojson/
extracted in step (2), sample_points.sh will generate a set
of 10 metre equidistant points along each road, storing the result in
data/csv_output. For convenience, the resulting sample points are also
aggregated into a single file data/csv_output/final/england_and_wales.csv.

Output

The result is a csv file for each city stored in
data/csv_output/CITY_NAME.csv. For example:

Where:

Road network classifications

Licence

Open Government license v3 OGL v3.