This is a graphical user interface tool for processing 3D kinect scan files. It can automatically remove background points from 3D Kinect scans, create watertight body shape surfaces, and visualize final noise-free meshes, providing data to generate realistic body shape and dimension predictions.
It works on macOS Mojave 10.14 or higher.
- Vertex/Actor selection
- Merge Meshes
- File Explorer
- Multithreading
$ python3 main.py
In 2014, the Bioscience Group at the University of Michigan Transportation Research Institute (UMTRI) developed a body shape and dimensions measurement method that can generate watertight realistic body shape along with predicted joint locations and anthropometric variables from a 3D scan of an individual (Park et al. 2014). Although this method works with the noisy point cloud data, it requires high computational power and is error-prone due to the noise and unnecessary scan points. The purpose of the current study is to develop an intuitive graphical user interface software that (1) automatically removes background points, (2) creates watertight surface from oriented point sets using Poisson surface reconstruction method and (3) visualizes final noise-free meshes. Using results generated by the program will improve the fitting performance and the body shape and dimension predictions.
- Create noise-free meshes to improve the fitting performance and the accuracy of the established body shape and dimensions measurement method
- Visualize results in 3D using a graphical user interface
- The user should be able to work through multiple projects one at a time, and set custom configuration for individual meshes