Methods for Feature Detection in Point Clouds

Christopher Weber (1) , Stefanie Hahmann (2) , Hans Hagen (1)

1 Department of Computer Science [Kaiserslautern]
2 EVASION - Virtual environments for animation and image synthesis of natural objects

This paper gives an overview over several techniques for detection of features, and in particular sharp features, on point-sampled geometry. In addition, a new technique using the Gauss map is shown. Given an unstructured point cloud, this method computes a Gauss map clustering on local neighborhoods in order to discard all points that are unlikely to belong to a sharp feature. A single parameter is used in this stage to control the sensitivity of the feature detection.

Mots clés: point cloud; sharp features; reconstruction; Gaussmap; clustering

Domaines: Synthèse d'image et réalité virtuelle [cs.GR]

Christopher Weber, Stefanie Hahmann, Hans Hagen. Methods for Feature Detection in Point Clouds. Visualization of Large and Unstructured Data Sets - Applications in Geospatial Planning, Modeling and Engineering (IRTG 1131 Workshop), Mar 2010, Bodega Bay, CA, United States. pp.90-99, ⟨10.4230/OASIcs.VLUDS.2010.90⟩. ⟨hal-00921790⟩