Pararoof:基于 Python 的点云屋顶参数化重建项目

Implementation of Para-Roof: High-Quality Parametric Roof Reconstruction via Primitive Proposal Extraction from Point Cloud

Branch1Tags0

Pararoof

Implementation of Para-Roof: High-Quality Parametric Roof Reconstruction via Primitive Proposal Extraction from Point Cloud

Requirements


  • python 3.9

Installation


Manually install the required dependencies using the following command:

pip install open3d
pip install numpy
pip install scikit-learn
pip install alphashape
pip install trimesh
pip install natsort
pip install tqdm
pip install pyqt5
conda install -c conda-forge pythonocc-core=7.8.1

Running the code


Parameters (If needed, you can manually modify these parameters.):

The parameter 'alpha' is the angle threshold for determining constraint imposition, with a default value of 5.
The parameter 'beta' is the angle threshold for determining when to merge coplanar planes, with a default value of 10.
The parameter 'mu_d' is the threshold for determining the primary normal vector, with a default value of the total number of points divided by 20.

Put the planar detection results(vertex groups (.vg)) in the input_vg folder and the mesh model corresponding to the input roof point cloud in the input_mesh folder. Use the following command to refine the roof planes and extract proposals:

python primitives_ex.py

The extracted proposals are saved in the .pkl format which are in the proposals_pkl folder.

Then use the following command to reconstruct the roof:

python cut_shape.py

The reconstruction results are saved in the roof_recon folder.

Introduction

Implementation of Para-Roof: High-Quality Parametric Roof Reconstruction via Primitive Proposal Extraction from Point Cloud

Customize your domain