Automated urban remote sensing with Python

Setup

Python environment

# Create a new virtual environment
conda create --name remotesensing python
source activate remotesensing

# Install packages
conda install numpy scipy scikit-learn matplotlib \
              pandas psycopg2 rasterio fiona \
              jupyter

Run the notebook

# Clone the repository
git clone https://github.com/yannforget/foss4g.git your_directory
cd your_directory

# Start jupyter notebook
jupyter notebook

Populate a local PostGIS database with OSM data

Using osm2pgsql:

# Start postgres if needed
sudo systemctl start postgres.service

# Login using psql
psql -U postgres

CREATE DATABASE your_database;
\connect your_database;
CREATE EXTENSION postgis

# Populate the database
osqm2pgsql --slim --create --multi-geometry --prefix osm \
  --database your_database --proj 32630 data/ouagadougou.xml