Title :
Unsupervised hierarchical fingerprint matching
Author :
Ozbayoglu, A. Murat ; Degli, C.H.
Author_Institution :
Smart Eng. Syst. Lab., Missouri Univ., Rolla, MO, USA
Abstract :
A hierarchial classification and identification model for fingerprint matching is presented. The proposed system categorizes the fingerprint into initial classes. The model uses enhanced fingerprints obtained by frequency domain processing as input, and classifies them with an unsupervised neural network. Finally, the identification is performed within these classes using another neural network with different features. The model is tested for several fingerprints, the result are analyzed and discussed
Keywords :
fingerprint identification; frequency-domain analysis; image classification; neural nets; frequency domain processing; hierarchial classification; identification model; unsupervised hierarchical fingerprint matching; unsupervised neural network; Band pass filters; Data mining; Feature extraction; Fingerprint recognition; Frequency; Image converters; Image matching; Impedance matching; Spatial databases;
Conference_Titel :
Neural Networks,1997., International Conference on
Conference_Location :
Houston, TX
Print_ISBN :
0-7803-4122-8
DOI :
10.1109/ICNN.1997.614006
