Setup | Usage | Data Sources | Code Organization | Citation | Acknowledgement | License
This repository provides the methodology and analysis for the paper "Measuring OpenStreetMap building footprint completeness using human settlement layers".
Geographic data is essential in supporting efforts for disaster risk reduction and preparedness. The goal of this project is to find the unmapped areas in OpenStreetMap so that more data can be contributed there. This will help OSM volunteers make decisions on which areas to map by letting them monitor and compare the completeness in different areas before they focus their efforts on mapping. In this study, we use human settlements data to measure the data completeness of OSM building footprints.
We also created an interactive web map and we wrote a blog to show the results in the Philippines.
Create a Python virtual environment and install the dependencies found in requirements.txt.
make venv
python3 -m venv venv
venv\Scripts\activate
pip install pip-tools
pip install -r requirements.txt
- Activate your virtual environment
source venv/bin/activate
venv\Scripts\activate
- Run JupyterLab
jupyter lab
We have three sources of GIS data for the Philippines and Madagascar.
- High Resolution Settlement Layer (HRSL) - high resolution population dataset from Facebook Data for Good
- Administrative Boundaries - administrative division/unit/boundary from Humanitarian Data Exchange (HDX)
- OpenStreetMap (OSM) - free wiki world map
Upon cloning, the repository contains the following directories:
- notebooks/ - has all Jupyter notebook for methodology and analysis
- 1_Methodology.ipynb - shows the methodology of the project: importing the packages, downloading the datasets
- 2_Analysis.ipynb - shows the analysis of the project: importing the packages, identifying urban or rural unmapped areas
- assets/ - images used for the readme
After running the notebooks from top-to-bottom and in order, the repository should have these new directories:
- download_data/ - raw and intermediately processed datasets
- data/ - processed datasets for analysis
- plots/ - plots from analysis
You can use this bibtex entry to cite this repository:
@misc{osm_completeness_2020,
title={Measuring OpenStreetMap building footprint completeness using human settlement layers},
author={Orden, Ardie and Flores, Ren Avell and Faustino, Pia and Samson, Mark Steve},
year={2020},
publisher={GitHub},
journal={GitHub Repository},
howpublished={\url{https://github.com/thinkingmachines/osm-completeness}},
}
We'd like to thank Nick Brown from Humanitarian OpenStreetMap Team (HOT) and Mikel Maron from the Mapbox community team for their feedback and support.
We'd also like to thank the OpenStreetMap Philippines community for their support and all their contributions to the map.
MIT Licence
Copyright 2020 Thinking Machines Data Science