Computing elevation maxima by searching the gauss sphere

  • Authors:
  • Bei Wang;Herbert Edelsbrunner;Dmitriy Morozov

  • Affiliations:
  • SCI Institute, University of Utah;Duke University, Geomagic and IST Austria;Stanford University

  • Venue:
  • Journal of Experimental Algorithmics (JEA)
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

The elevation function on a smoothly embedded 2-manifold in ℝ3 reflects the multiscale topography of cavities and protrusions as local maxima. The function has been useful in identifying coarse docking configurations for protein pairs. Transporting the concept from the smooth to the piecewise linear category, this article describes an algorithm for finding all local maxima. While its worst-case running time is the same as of the algorithm used in prior work, its performance in practice is orders of magnitudes superior. We cast light on this improvement by relating the running time to the total absolute Gaussian curvature of the 2-manifold.