Izvorni znanstveni članak
FINGERPRINTS PREPROCESSING USING WALSH FUNCTIONS
Miroslav Bača
; Fakultet organizacije i informatike Sveučilišta u Zagrebu, Varaždin, Hrvatska
Kornelije Rabuzin
orcid.org/0000-0002-0247-669X
; Fakultet organizacije i informatike Sveučilišta u Zagrebu, Varaždin, Hrvatska
Zvonko Merkaš
orcid.org/0000-0003-1770-3952
; Autohrvatska, Zagreb, Hrvatska
Sažetak
Minutiae classification and fingerprint classification in fingerprint evaluating process are very important. Fingerprint image contains about 150 minutiae’s. When we compare two fingerprint images, we compare latent and non latent fingerprint and we try to find 12 minutiae’s placed on the same position on latent and non latent fingerprint images.
After fingerprint image pre-processing we can perform classification or we can try to find minutiae. In this paper we describe the process of minutiae classification for comparison purposes. For that purpose we use Walsh function and Walsh transforms. Paper describes minutiae classification which is relatively new in recognition systems and gives contribution for development of practical fingerprint recognition systems. Paper also gives contribution in the theoretical part due to the fact that Walsh functions were not implemented in fingerprint pre-processing systems so far. The new symbolic database model for fingerprint storage gives multifunctional foundations for future research.
Ključne riječi
fingerprint; minutiae; Walsh transform; classification
Hrčak ID:
20840
URI
Datum izdavanja:
10.6.2006.
Posjeta: 2.915 *