Ifc to cloudpoint semi synthetic data generator
This project is a collection of scripts to generate point clouds from IFC files.
The point clouds can be used for machine learning and other applications.
The synthetic data is generated by extracting the meshes by ifc-entity-type and distributing points over each face using barycentric coordinates and a weightend sampler-algorithm with respect to triangle-size.
All mesh-points are concatenated to a single point cloud.
The point cloud is then saved as a .ply file with the format float float float int, where the first three values are coordinates and the fourth value is the class label.
Currently a DGCNN model is implemented to segment the point clouds.
The model is trained on 13 ifc types, to segment the point clouds into 13 classes.
The dataset used is very small (9) and split into train (7) and test (2) set.
git clone
cd ifcclouds
python3 -m venv venv
source venv/bin/activate
python3 -m pip install -r requirements.txt
Download the IFC files and extract them to the data/raw folder. The folder structure should look like this:
data
└── raw
├── sample1.ifc
├── sample2.ifc
├── sample3.ifc
├── sample4.ifc
├── sample5.ifc
├── sample6.ifc
├── sample7.ifc
├── sample8.ifc
...
python3 -m ifcclouds.data.make_dataset
python3 -m ifcclouds.convert IFC_FILE_PATH PLY_FILE_PATH
or in script
from ifcclouds.convert import process_ifc
python3 -m ifcclouds.server.app
- Improve dataset consumption, to distribute well over classes
- Add more models
├── LICENSE
├── Makefile <- Makefile with commands like `make data` or `make train`
├── README.md <- The top-level README for developers using this project.
├── data
│ ├── external <- Data from third party sources.
│ ├── interim <- Intermediate data that has been transformed.
│ ├── processed <- The final, canonical data sets for modeling.
│ └── raw <- The original, immutable ifc data dump.
│
├── docs <- A default Sphinx project; see sphinx-doc.org for details
│
├── models <- Trained and serialized models, model predictions, or model summaries
│
├── notebooks <- Jupyter notebooks. Naming convention is a number (for ordering),
│ the creator's initials, and a short `-` delimited description, e.g.
│ `1.0-jqp-initial-data-exploration`.
│
├── references <- Data dictionaries, manuals, and all other explanatory materials.
│
├── reports <- Generated analysis as HTML, PDF, LaTeX, etc.
│ └── figures <- Generated graphics and figures to be used in reporting
│
├── requirements.txt <- The requirements file for reproducing the analysis environment, e.g.
│ generated with `pip freeze > requirements.txt`
│
├── setup.py <- makes project pip installable (pip install -e .) so src can be imported
├── src <- Source code for use in this project.
│ ├── __init__.py <- Makes src a Python module
│ │
│ ├── data <- Scripts to download or generate data
│ │ └── make_dataset.py
│ │
│ ├── features <- Scripts to turn raw data into features for modeling
│ │ └── build_features.py
│ │
│ ├── models <- Scripts to train models and then use trained models to make
│ │ │ predictions
│ │ ├── predict_model.py
│ │ └── train_model.py
│ │
│ └── visualization <- Scripts to create exploratory and results oriented visualizations
│ └── visualize.py
│
└── tox.ini <- tox file with settings for running tox; see tox.readthedocs.io
Project based on the cookiecutter data science project template. #cookiecutterdatascience