Anisotropic semi-pseudometrics

Article type: Research Article

Authors: Lowen, R. | Peeters, W.

Affiliations: RUCA University of Antwerp, Groenenborgerlaan 171, B-2020 Antwerpen, Belgium. Tel.: +32 3 218 02 92; Fax: +32 3 218 02 04; E-mail: {lowen, wpeeters}@ruca.ua.ac.be

Abstract: The purpose of this article is to describe a recognition system, used in the matching of patterns that are distorted by various kinds of noise. The proposed technique is more general than the one described in [12], but also makes fundamental use of the concept of semi-pseudometrics. It will be shown that in some cases, the chance of correctly recognizing an image increases when using these in a distance classifier.

Journal: Journal of Intelligent and Fuzzy Systems, vol. 8, no. 2, pp. 147-158, 2000