Improving the Ridge Based Fingerprint Recognition Method Using Sweat Pores

Among several biometric traits possible to be used for people identification, fingerprint is still the most used. Current automated fingerprint identification systems are based on ridge patterns and minutiae, classified as first and second level features, respectively. However, the development of new fingerprint sensors and the growing demand for more secure systems are leading to the use of additional discriminative fingerprint characteristics known as third level features, such as the sweat pores. Recent researches on fingerprint recognition have focused on fingerprint fragments, in which methods based only on first and second level features tend to obtain low recognition rates. This paper proposes a robust method developed for fingerprint recognition from fingerprint fragments based on ridges and sweat pores. We have extended a ridgebased fingerprint recognition method previously proposed in the literature, based on Hough Transform, by incorporating sweat pores information in the matching step. Experimental results showed that although the reduction of Equal Error Rate is modest, a significant improvement was observed when analyzing the FMR100 and FMR1000 metrics, which are more suitable for high security applications. For these two metrics, the proposed approach obtained a reduction superior to 10% of the rates, when compared to the original ridge-based approach. Keywords-biometrics; fingerprints; ridges; sweat pores