Fig 2 - uploaded by Asker M. Bazen
Content may be subject to copyright.
Prefix code implementation by means of a node list.
Source publication
In this paper, an environment for using genetic programming is presented. Although not restricted to a spe-cific domain, our intention is to apply it to image processing problems such as fingerprint recognition. The environment performs tasks like: population management, genetic oper-ators and distributed parallel evaluation of the programs. Furthe...
Context in source publication
Citations
... It turns out that the traditional approach provides the best quality minutiae extraction. Part of the traditional approach has been published in [Baz02a], the genetic programming approach has been published in [Meu00,Meu01] and the reinforcement learning approach has been published in [Baz01e]. ...
Presents an overview of state-of-the-art fingerprint recognition technology for identification and verification purposes. Three principal challenges in fingerprint recognition are identified: extracting robust features from low-quality fingerprints, matching elastically deformed fingerprints and efficiently searching through a fingerprint database. A number of algorithms are proposed that increase the recognition performance, including new algorithms for feature extraction, and an algorithm that explicitly compensates for non-linear deformations in the fingerprint images. An indexing method that reduces the processing time and error rates for fingerprint searches involving large databases is also presented. Applications of these algorithms in fingerprint recognition systems will promote a more widespread use of fingerprint recognition in commercial applications. Krantenartikel uit de Telegraaf van 31-08-2002: BETROUWBAARDER EN HONDERD KEER SNELLER : HERKENNING AFDRUK VAN VINGER VERFIJND Betalen met een vingerafdruk, gebruik van mobiele telefoons met een vingerafdruk en entree tot gebouwen via een vingerafdruk. Als het aan promovendus ir. Asker Bazen ligt, zijn deze toepassingen zeer nabij. Bazen verbeterde de techniek om vingerafdrukken automatisch te herkennen met een 'elastische' methode. Hiermee is ook herkenning mogelijk van vingerafdrukken die zijn vervormd of beschadigd. Verder gaat de herkenning bij grote databases ongeveer honderd keer sneller dan voorheen. De elektrotechnicus, wiens bevindingen internationaal in hoog aanzien staan, promoveert op 18 september aan de Universiteit Twente. Bazen stelt dat de hogere snelheid en betrouwbaarheid drempels kunnen wegnemen voor acceptatie van de techniek. De vingerafdruk is volgens hem van alle biometrische herkenningsmethoden de meest praktische: te herkennen met een simpele sensor die op een slimme manier is gekoppeld aan een database. "Een sensor en bijbehorende intelligentie inbouwen in een compact apparaat zoals een mobiele telefoon is straks geen probleem meer", aldus Bazen. Dit is niet het geval bij bijvoorbeeld irisdetectie. Ook een zeer betrouwbare vorm van biometrie, maar bij deze methode is wel een goede camera nodig. Justitie neemt een proef om te kijken of identificatie van medewerkers veiliger kan.
... This paper will present the considerations related to an efficient implementation of such an environment as well as experimental results. Our GP environment, which is also briefly described in [6] , is called Poor Man's Distributed Genetic Programming (PMDGP), as it is designed to make use of existing heterogeneous hardware rather than expensive hardware that has to be purchased for the purpose of GP. The rest of this paper is structured as follows. ...
In this paper, an environment for using genetic pro-gramming is presented. Although not restricted to a specific domain, our intention is to apply it to im-age processing problems such as fingerprint recogni-tion. The environment performs tasks like: population management, genetic operators and distributed par-allel evaluation of the programs. Furthermore, it pro-vides a framework for implementation of the problem-specific part. Using object-oriented methods, the en-vironment is designed to offer a high degree of flexi-bility and ease of use. The GP environment uses distributed fitness eval-uation, which can be used on existing computer net-works. The system is optimized for efficiency of dis-tribution. Experiments are included in the paper to il-lustrate the high performance of our implementation of this distribution method.
