|
|
Robust Surface Reconstruction from Noisy Point Clouds |
|---|---|
| Objective | Implement an algorithm that is able to robustly extract a triangle mesh from a point cloud obtained from multiple depth maps that may contain outliers. |
| Description | 
Overview - Fuse all individual depth maps building a Delaunay triangulation of the 3D point set - Extract the surface of the scene from the triangulation based on visibility / photo-consistency using graph cut optimization - Variational mesh refinement -Compare the results with a standard Poisson surface reconstruction Keywords: 3D Computer Vision, Visualization, 3D Delaunay Triangulation, Graph Cuts Requirements: - C++, CGAL, Linux |
| Team | Computer Vision |
| Contact | Rumpler Markus |
| Offered as |
Semester project Master's thesis |
| Duration |
|
