Updated Mar 23, 2024
This repo develops an easy-to-use Python package, named OSMsc, to improve the availability, consistency and generalizability of urban semantic data.
OSMsc v0.2.0 is coming!
Photo by Abigail Keenan on Unsplash
- Construct semantic city objects based on the public dataset (OpenStreetMap), and apply geometric operations to build more complete city objects;
- Fuse 3D and tag information from multiple data sources through the spatial analysis between OSMsc layers and other non-OSM data layers;
- Propose the semantic connector(UrbanTile), and supplement the spatial semantics;
- Output the CityJSON-formatted semantic city models.
- 内部集成了OSM数据的自动化下载,仅需几行简单的代码,即可完成城市对象的构建;
- 轻松融合外部3D或者文本数据,丰富OSM城市对象的信息;
- 城市模型内部的对象可以添加空间语义,彼此之间的空间关系可以查询或者推测出来;
- 可以输出CityJSON或者html的可视化文件,CityJSON格式文件可以由https://ninja.cityjson.org 查看.
OSMsc workflow
Semantic city model generated by OSMsc
Install from Github
git clone https://github.com/ruirzma/osmsc.git
cd osmsc/
pip install .
or python setup.py install
Install from PyPi
pip install osmsc
Note:
-
OSMnx should be installed before OSMsc, installation errors of OSMnx could be resolved in the latest OSMnx documentation.
-
If installing OSMnx manually, pls download the Python extension packages (Rtree, GDAL, Fiona, rasterio, etc.) from here for Windows and Homebrew🍺 for MacOS.
If you use OSMsc in scientific work, I kindly ask you to cite it:
@article{doi:10.1080/13658816.2023.2266824,
author = {Rui Ma, Jiayu Chen, Chendi Yang and Xin Li},
title = {OSMsc: a framework for semantic 3D city modeling using OpenStreetMap},
journal = {International Journal of Geographical Information Science},
volume = {0},
number = {0},
pages = {1-26},
year = {2023},
publisher = {Taylor & Francis},
doi = {10.1080/13658816.2023.2266824},
URL = {https://doi.org/10.1080/13658816.2023.2266824},
eprint = {https://doi.org/10.1080/13658816.2023.2266824}}
1. Geoff Boeing, OSMnx, https://github.com/gboeing/osmnx
2. Nick Bristow, OSMuf, https://github.com/AtelierLibre/osmuf
3. Joris Van den Bossche, GeoPandas, https://github.com/geopandas/geopandas