2D vectorization project

  • T: now working on SGP re-submission, Sec. 4
  • the figure with junctions will replace current limitation figure with the high-valence junction
  • high valence junctions: we can say that with a proper frame field (with automatically-detected or user-specified singularities), even high valence junctions could be processed properly. We can also remark that high valence junctions are quite rare in practice.

3D curves project

  • T: results starting to look good
  • D: we could compare our method to a baseline = alpha shapes + medial axis. This could work well for simple cases, but should fail for more complex ones.
  • A: this is analogous to 2D skeletonization methods
  • CGAL: 3D alpha shapes
  • CGAL: Skeletonization

Stuff