Regularization and Scale Space

Mads Nielsen 1 Luc Florack 1 Rachid Deriche 1
1 ROBOTVIS - Computer Vision and Robotics
CRISAM - Inria Sophia Antipolis - Méditerranée
Abstract : Computational vision often needs to deal with derivatives of digital images. Derivatives are not intrinsic properties of a digital image; a paradigm is required to make them well-defined. Normally, a linear filtering is applied. This can be formulated in terms of scale space, functional minimization or edge detection filters. In this paper, we take regularization (or functional minimization) as a starting point, and show that it boils down to a ordered set of linear filters of which the Gaussian is the first if we require the semi group constraint to be fulfilled. This regularization implies the minimization of a functional which contains terms up to infinite order of differentiation. If the functional is truncated at second order, the Canny-Deriche filter arises. Furthermore, we show that the $n$th order Canny-optimal edge detection filter implements $n$th order regularization. We also show, that higher dimensional regularization in its most general form boils down to a rotation of the one dimensional case, when Cartesian invariance is imposed. This means that results from 1D regularization are easily generalized to higher dimensions. Finally, we show that regularization in its most general form can be implemented as recursive filtering without any approximation.
Type de document :
[Research Report] RR-2352, INRIA. 1994
Liste complète des métadonnées
Contributeur : Rapport de Recherche Inria <>
Soumis le : mercredi 24 mai 2006 - 15:04:00
Dernière modification le : samedi 27 janvier 2018 - 01:31:02
Document(s) archivé(s) le : mardi 12 avril 2011 - 16:32:49



  • HAL Id : inria-00074325, version 1



Mads Nielsen, Luc Florack, Rachid Deriche. Regularization and Scale Space. [Research Report] RR-2352, INRIA. 1994. 〈inria-00074325〉



Consultations de la notice


Téléchargements de fichiers