Surface reconstruction through poisson disk sampling.

Area of Science:

• Computational Geometry
• Computer Graphics
• Mesh Generation

Background:

• Generating Voronoi diagrams for large, scattered point sets can be computationally intensive.
• Redundant points in datasets can lead to inefficiencies in mesh model construction.
• Geodesic distances are crucial for accurate representations on complex surfaces.

Purpose of the Study:

• To develop an efficient method for generating approximate Voronoi diagrams in the geodesic metric.
• To address the challenge of redundant scattered points in data.
• To construct accurate mesh models from selected sample points.

Main Methods:

• Utilized Poisson disk sampling to select representative seed points from original data.
• Generated an approximate Voronoi diagram based on geodesic distances among selected seeds.
• Constructed a mesh model using the generated Voronoi diagram.

Main Results:

• The proposed method effectively handles redundant scattered points by sampling.
• The generated mesh models reflect the level of detail of the original point set.
• Experimental evaluations demonstrate the reliability and effectiveness of the approach.

Conclusions:

• The strategy of generating approximate geodesic Voronoi diagrams using sampled points is efficient.
• This method provides a reliable way to construct mesh models from dense point clouds.
• Adaptive reconstruction is achievable by adjusting seed selection strategies.