A parallel algorithm for computing the flow complex

  • Authors:
  • Joachim Giesen;Lars Kuehne

  • Affiliations:
  • University Jena, Jena, Germany;University Jena, Jena, Germany

  • Venue:
  • Proceedings of the twenty-ninth annual symposium on Computational geometry
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present a parallel algorithm and its implementation for computing the entire Hasse diagram of the flow complex of a point cloud in Euclidean space. Known algorithms for computing the flow complex in two and three dimensions compute the geometric realization of the flow complex and need to compute the Delaunay triangulation of the point cloud first. Our algorithm computes less information, namely only the Hasse diagram of the flow complex that is augmented with enough geometric information to allow the same topological multi-scale analysis of point cloud data as the alpha shape filtration without computing the Delaunay triangulation explicitly. We show experimental results for medium dimensions that demonstrate that our algorithm scales well with the number of available cores on a multicore architecture.