Fingerprint Recognition System for Personal Identification

Fingerprints are widely used in biometrictechniques for automatic personal identificationwork. In this paper, a fingerprint recognition systemis developed to identify a person according tofingerprint image on Myanmar National RegistrationCards (NRCs). Generally, fingerprint identificationapproaches are minutiae-based and texture-based.Although the minutiae-based method is popular andextensively used method for fingerprintidentification, it shows poor performance for lowquality images. In proposed system, the texturebasedapproach for fingerprint recognition usingDiscrete Wavelet Transform (DWT) is developed. Toreduce the search time and computationalcomplexity, the proposed system classifies thefingerprint types according to ridges direction bysingular candidate analysis using an extendedrelational graph. And then, the system finds the localfeatures of the fingerprint using DWT and it iscompared to the subset of the database containingthat type of fingerprints using Euclidean distancemetric. The performance of the proposed system canbe evaluated by measuring its False Reject Rate(FRR) and False Accept Rate (FAR). Theeffectiveness of the proposed system can beconfirmed through the experimental results