B-spline-Based surface reconstruction for the analysis of distorted panels

Kuvaus

Application for the fitting of B-spline surfaces to point clouds. The motivation is to provide an easy and accesible programm to characterize distortion shapes of welded thin-deck ship panels to enable further investigations on their strength capabilities. Table of Contents Installation Dependencies Usage Installation The scripts can be used in two ways; either as standalone scripts, where the raw functions are provided or as an application to enable easy usability. Please make sure, that the environment is setup in Python 3.8. Dependencies Dependency check is implemented in the beginning of each script, with the option to install during execution. Alternatively, these libraries can be easily installed through pip install Open3D: Library for 3D data processing. Used for voxel downsampling and visualization in this project. numpy: Fundamental package for scientific computing in Python. meshlab: pyMeshLab is a Python library that enables users to write scripts using the various functions available in Meshlab. Furthermore, the installation of FreeCAD v0.21 on the local machine is required. The local installation path needs to be added to the script `surface_fitting.py` in the following line. Replace USER with the user on the local machine. # insert path here that goes to the installation of FreeCAD FREECADPATH = r"C:\\Users\\USER\\AppData\\Local\\Programs\\FreeCAD 0.21\\bin" FREECADLIBPATH = r"C:\\Users\\USER\\AppData\\Local\\Programs\\FreeCAD 0.21\\lib" Usage The package can be used either by running `ui2_app.py` which gives an GUI for the user to setup the surface reconstruction process. The second option is to utilize the scripts `crop_geometry.py` `voxel_downsampling.py` and `surface_fitting.py` to build an own application 1. Run the Application: python ui2_app.py 2. Follow the GUI instructions: Starting point for the reconstruction is a planar type point cloud in `.ply` format. Load the `.ply` file. Set up parameters for the surface reconstruction process. Parameters are strongly case dependent, recommendations for plate fields (based on investigations done in the paper) are: voxel size = 2 UDegree = 3 VDegree = 3 NbUPoles = 15 NbVPoles = 15 Iterations = 3 3. Perform reconstruction and error analysis: The reconstructed surface will be saved as `.step` format in the specified output folder. The application also gives maximum / minimum point to surface distance, as well as the mean average error of the reconstruction as output per default. These can be changed by modifying `error_estimation.py`.
Näytä enemmän

Julkaisuvuosi

2025

Aineiston tyyppi

Tekijät

Department of Energy and Mechanical Engineering

Federica Mancini Orcid -palvelun logo - Muu tekijä

Heikki Remes Orcid -palvelun logo - Muu tekijä

Matti Christmann - Tekijä

Bauhaus-Universität Weimar - Muu tekijä

Lappeenranta University of Technology - Muu tekijä

Zenodo - Julkaisija

Projekti

Muut tiedot

Tieteenalat

Kone- ja valmistustekniikka

Kieli

Saatavuus

Avoin

Lisenssi

Creative Commons Nimeä 4.0 Kansainvälinen (CC BY 4.0)

Avainsanat

Asiasanat

Ajallinen kattavuus

undefined

Liittyvät aineistot