Particle-based sampling and meshing of surfaces in multimaterial volumes

Area of Science:

• Computer Graphics
• Computational Geometry
• Scientific Visualization

Background:

• Accurate geometric capture of complex material interfaces in volume data is crucial for realistic simulations and visualizations.
• Existing methods face challenges in generating well-sampled, consistent tessellations and respecting non-manifold geometry at material boundaries.

Purpose of the Study:

• To develop a robust strategy for sampling and meshing multimaterial volumes from labeled data.
• To address the challenges of consistent tessellation and non-manifold geometry in multimaterial interfaces.

Main Methods:

• Utilizes dynamic particle systems for sampling multimaterial volumes.
• Introduces a novel, differentiable representation for material junctions to explicitly sample intersections.
• Employs fundamental sampling constraints to guide particle distribution.
• Applies Delaunay-based meshing algorithms to extract watertight meshes.

Main Results:

• The proposed method successfully generates high-quality, watertight meshes for complex multimaterial interfaces.
• The differentiable representation enables explicit sampling of corners, edges, and surfaces of material intersections.
• Particle distribution control ensures reliable mesh extraction by Delaunay algorithms.

Conclusions:

• The dynamic particle system approach offers a robust solution for meshing multimaterial volumes.
• This method enhances the accuracy and realism of visualizations and simulations involving complex geometries.
• The technique effectively handles non-manifold geometry and ensures mesh consistency across material boundaries.