Biometric identification still remains one of the most reliable and consistent methodologies for identification as established over the years from various studies among other authentication systems. Fingerprint biometrics is the most commonly and universally accepted biometric due to its uniqueness among all other biometrics. Minutiae algorithm remains the widely and popularly used method but still needs improvement for better performance due to its poor performance in real-time validation. The problems of minutiae method can be eliminated and its performance enhanced by combining correlation approach. This paper used a hybrid fingerprint matching system with the fusion of minutiae and correlation features. The final matching score is calculated by fusing matching scores of the two algorithms. The False Acceptance Rate (FAR), False Recognition Rate (FRR), and the Mean Evaluation Metrics (MEM) were used for performance evaluation of the proposed system. The hybridized system achieves better results when compared with other related works on commonly used databases Fingerprint Verification Competition (FVC) 2002. The authors obtained FAR 0.0038, FRR 0.0156 and an accuracy 99.25%. The evaluation results show an effective performance over the use of only minutiae algorithm for identification and authentication. Moreso, the results show the improvement in performance of the proposed method and the added advantages of combining both algorithms for identifications systems and performed better in real time identification and verification of large databases.
Keywords: Hybrid system, correlation technique, fingerprint identification, Minutiae algorithm, biometric.