Related Experiment Video
Updated: Jul 6, 2025
11:06
3D Printing of Preclinical X-ray Computed Tomographic Data Sets
Published on: March 22, 2013
40.5K
Visual-Preserving Mesh Repair
IEEE Transactions on Visualization and Computer Graphics
|January 1, 2024
Summary
This study presents a novel mesh repair framework using visual measures to fix defective 3D models. The method effectively closes gaps, reorients faces, and optimizes meshes while preserving visual quality and attributes for various applications.
Area of Science:
- Computer Graphics
- Geometric Modeling
Background:
- Mesh repair is crucial for downstream applications like simulation and fabrication.
- Defective meshes with gaps, holes, and orientation inconsistencies pose significant challenges.
Purpose of the Study:
- To develop a robust mesh repair framework that addresses common mesh defects.
- To preserve visual appearance and geometric accuracy during the repair process.
- To reduce mesh complexity without compromising quality for rendering and further processing.
Main Methods:
- Introduced three visual measures (visibility, orientation, openness) based on ray-tracing.
- Developed a novel framework incorporating these measures for open surface closing, face reorientation, and global optimization.
- Preserved input attributes like UV coordinates for rendering.
Main Results:
- Effectively repaired meshes with various defects including gaps, holes, self-intersections, and inconsistent orientations.
- Reduced mesh complexity while maintaining geometric accuracy and visual quality.
- Demonstrated robustness and effectiveness on hundreds of models from ShapeNet and Thingi10K.
Conclusions:
- The proposed mesh repair framework offers an effective solution for converting defective meshes into watertight manifold meshes.
- The method successfully preserves visual appearance and essential attributes, benefiting diverse computer graphics applications.
- The approach is robust and outperforms existing methods in mesh repair tasks.
