Fingerprint recognition using Gabor filters and Wavelet features [articol]

Abstract

Amongst the several approaches on fingerprint matching, the most popular continues to be minutiae based. But this technique does not use much of the discriminatory information available in the fingerprints, because local ridge structures cannot be completely characterized by minutiae, further more these methods are slow in comparing fingerprint images containing different number of minutiae. Our method uses a combined structure of Gabor filters and Wavelet features, in order to obtain a well balanced aggregated decision. The recognition system is reinforced in a biometric approach, using in parallel a speaker recognition module. The final decision is obtained by using Dempster - Schafer evidence theory in a hierarchic structure.