3d-building-metrics
3d-building-metrics copied to clipboard
tudelft3d
Tool to compute shape metrics from 3D geometries for buildings
3DBM
3D Building Metrics. Elevating geometric analysis for urban morphology, solar potential, CFD etc to the next level 😉
Installation
You need to install all dependencies first:
pip install -r requirements.txt
Then take your time and install pymesh.
Wat is het?
A cool script that computes a lot metrics from 3D geometries (mostly intended for buildings).
The following metrics are computed:
| Type | Metrics |
|---|---|
| Geometric Properties | Number of vertices, Number of surfaces, Number of vertices by semantic type (i.e. ground, roof, wall), Number of surfaces by semantic type (i.e. ground, roof, wall), Min/Max/Range/Mean/Median/Std/Mode height |
| Derived Properties | Footprint perimeter, Volume, Volume of convex hull, Volume of Object-Oriented Bounding Box, Volume of Axis-Oriented Bounding Box, Volume of voxelised building, Length and width of the Object-Oriented Bounding Box, Surface area, Surface area by semantic surface, Horizontal elongation, Min/Max vertical elongation, Form factor |
| Spatial distribution | Shared walls, Nearest neighbour |
| Shape indices | Circularity/Hemisphericality*, Convexity 2D/3D*, Fractality 2D/3D*, Rectangularity/Cuboidness*, Squareness/Cubeness*, Cohesion 2D/3D*, Proximity 2D/3D+, Exchange 2D/3D+, Spin 2D/3D+, Perimeter/Circumference*, Depth 2D/3D+, Girth 2D/3D+, Dispersion 2D/3Dx, Range 2D/3D*, Equivalent Rectangular/Cuboid*, Roughnessx |
- * formula-based index, size-independent by definition
- + index based on interior grid points (discretised), normalised
- x index based on surface grid points (discretised), normalised
Omg, how amazing! Any issues?
Yeah:
- It works with only
MultiSurfaceandSolid(the latter, only for the first shell) - It only parses the first geometry
- Expects semantic surfaces
How?
Running it, saving it, and including a val3dity report:
python cityStats.py [file_path] -o [output.csv] [-v val3dity_report.json]
Default is single-threaded, define the number of threads with:
python cityStats.py [file_path] -j [number]
Visualising a specific building, which can help with troubleshooting:
python cityStats.py [file_path] -p -f [unique_id]
Running multiple files in a folder and checking with val3dity (make sure you have val3dity installed):
for i in *.json; do val3dity $i --report "${i%.json}_v3.json"; python cityStats.py $i -o "${i%.json}.csv" -v "${i%.json}_v3.json"; done
Can I visualise a model?
Tuurlijk! Just:
python cityPlot.py [file_path]
Tutorial please!
-
Download or
git clonethis repository. -
Install all dependencies:
pip install -r requirements.txt. -
Download a tile from 3D BAG:
wget --header='Accept-Encoding: gzip' https://data.3dbag.nl/cityjson/v210908_fd2cee53/3dbag_v210908_fd2cee53_5910.json -
Run the stats on the data:
python cityStats.py 3dbag_v210908_fd2cee53_5910.json -o 5910.csv -
The resutling file
5910.csvcontains all metrics computed for this tile.
You may also run this with a val3dity report. You may download the val3dity report as a json file from the aforementioned website. Assuming the report's filename is report.json you can run:
python cityStats.py 3dbag_v210908_fd2cee53_5910.json -v report.json -o 5910.csv
Then the result will contain more info related to the validation of geometries.
If you use 3DBM in a scientific context, please cite this article:
Anna Labetski, Stelios Vitalis, Filip Biljecki, Ken Arroyo Ohori & Jantien Stoter (2023): 3D building metrics for urban morphology. International Journal of Geographical Information Science, 37(1): 36-67. DOI: 10.1080/13658816.2022.2103818
@article{Labetski2023,
Author = {Anna Labetski and Stelios Vitalis and Filip Biljecki and Ken {Arroyo Ohori} and Jantien Stoter},
Title = {{3D} building metrics for urban morphology},
Journal = {International Journal of Geographical Information Science},
Volume = {37},
Number = {1},
Pages = {36-67},
Year = {2023},
Publisher = {Taylor & Francis},
Doi = {10.1080/13658816.2022.2103818}
tudelft3d