Visibility-driven mesh analysis and visualization through graph cuts.

Area of Science:

• Computer Graphics
• Computational Geometry
• Geometric Modeling

Background:

• Accurate mesh processing is crucial for 3D data.
• Handling complex models with defects like holes and intersections remains challenging.

Purpose of the Study:

• To present a robust algorithm for classifying triangle faces as inside or outside.
• To demonstrate the algorithm's utility in practical 3D applications.
• To address limitations of existing methods when dealing with imperfect input meshes.

Main Methods:

• The algorithm operates directly on the input triangular mesh.
• It employs sampling to construct a visibility graph.
• Graph cut segmentation is utilized to classify faces.

Main Results:

• The algorithm robustly classifies triangle faces on complex models.
• Demonstrated success in normal orientation, inside removal, and layer-based visualization.
• Maintains accuracy despite input issues like holes, coplanar, and intersecting triangles.

Conclusions:

• The proposed algorithm offers a robust solution for inside/outside triangle classification.
• It effectively handles challenging 3D models, improving reliability in geometric applications.
• The method provides a foundation for advanced mesh processing tasks.