New Operator Based on a Multi Support Point Algorithm Applied to Feature Extraction - Archive ouverte HAL Access content directly
Conference Papers Year :

New Operator Based on a Multi Support Point Algorithm Applied to Feature Extraction

Abstract

In the context of quality control, we propose a new operator using multi support points to highlight local perturbations on flat surfaces. We compare the design of this operator with different gradients and residues in their ability to extract small perturbations as well as their efficiency on large surface. The operator is defined as a weighted differential operator evaluated on each point. It is designed to extract a region with a high slope followed by a plateau with a given width. It has a low computational complexity and it could be vectorized.
Fichier principal
Vignette du fichier
ics13Bilodeau.pdf (690.11 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-00880286 , version 1 (05-11-2013)

Identifiers

  • HAL Id : hal-00880286 , version 1

Cite

Michel Bilodeau, Estelle Parra-Denis, Dominique Jeulin. New Operator Based on a Multi Support Point Algorithm Applied to Feature Extraction. International Congress for Stereology, Oct 2011, Beijing, China. ⟨hal-00880286⟩
4657 View
34 Download

Share

Gmail Facebook Twitter LinkedIn More