A comparative study of various minutiae extraction methods for fingerprint recognition based on score level fusion

Abstract

A Multimodal Biometric system combines the evidences from various biometric sources or multiple evidences from single biometric source to atone for the limitations in performance of unimodal biometric system. This paper discusses two Minutiae extraction techniques to recognize fingerprint based on confidence level fusion of two extracted features, bifurcations and ridge endings and compares the recognition accuracy. In particular, the well known Morphological based minutiae extraction approach is compared with the proposed fuzzy logic control based approach. Experimental results based on IITD fingerprint database demonstrate that the score level fusion of bifurcations and ridge endings for fingerprint leads to a dramatically improvement in performance. And also the results reveal that our proposed fuzzy logic control based minutiae extraction is much more reliable than the Morphological based minutiae extraction approach.