Fingerprint matching using friction ridge counts

Abstract

As an alternative to current minutiae-based fingerprint matching techniques, a "circular" matching method is introduced. This method involves choosing the "best" center of the fingerprint, radiating circles outward from the center, and then recording the number of ridge crossings for each radius. From the vector of radii values and ridge crossing counts, a profile function is created. The profile functions of different fingerprints are compared using the absolute integrated difference. Box-Jenkins (ARIMA) time series models, and time series spectral analysis.