EER, Robustness and Generation Condition of Fingerprint Templates Based on the Fractional Cosine and Sine Transforms

Abstract

Recently, we have proposed new fingerprint templates based on the 1D discrete fractional Fourier, cosine and sine transforms (DFRT, DFCT and DFST) in order to realize the fingerprint recognition system with high recognition accuracy and high robustness against attacks. In this study, the generation condition of the fingerprint templates, i.e., a range of the transforms' orders, p i , is investigated in more detail to realize higher recognition accuracy. In addition, the EERs and the robustness are compared between the cases that a set of the transforms' orders is changed for each person and the set is unchanged for each person. As a result, in the case that the set is changed for each person, the most appropriate templates can be obtained as the phase distributions (PDs) of the DFCTs and the DFSTs under the condition that 0.0< p i ≤0.3 and 0.0< p i ≤0.5.The EER is an order of 10 -7 % and fully smaller than that (2.45%) of the original fingerprint images. On the other hand, in the case that the set is unchanged for each person, the PDs of the DFCTs is the most appropriate templates under the condition that 0.0< p i ≤0.3. The EER is 3.81% and a little bit higher than that of the original fingerprint images. The PSNRs are an order of several dB for both of the cases and it is found that proposed fingerprint templates have fully high robustness.