Geometric Feature Estimators for Noisy Discrete Surfaces

Laurent Provot (1) , Isabelle Debled-Rennesson (1)

1 ADAGIO - Applying Discrete Algorithms to Genomics and Imagery

We present in this paper robust geometric feature estimators on the border of a possibly noisy discrete object. We introduce the notion of patch centered at a point of this border. Thanks to a width parameter, attached to a patch, the noise on the border of the discrete object can be considered, and an extended flat neighborhood of a border point is computed. Stable geometric features are then extracted around this point. A normal vector estimator is proposed as well as a detector of convex and concave parts on the border of a discrete object.

Mots clés: Digital geometry; Geometric feature estimators; 3D discrete objects

Domaines: Mathématique discrète [cs.DM]

Laurent Provot, Isabelle Debled-Rennesson. Geometric Feature Estimators for Noisy Discrete Surfaces. 14th IAPR International Conference on Discrete Geometry for Computer Imagery - DGCI 2008, Apr 2008, Lyon, France. pp.275-286, ⟨10.1007/978-3-540-79126-3_25⟩. ⟨hal-00345284⟩