Fingerprint Biometric System Based on Minutiae Score Matching

Abstract

In recent days, biometrics plays an important role in forensic and military applications. In this paper, we proposed a biometric system using finger print. The captured finger print images are low quality in nature. These finger print images are enhanced using Contrast Limited Adaptive Histogram Equalization technique. Then the features are extracted from the finger print images. The minutiae features are extracted and encrypted using secret sharing based cryptographic algorithm. The encrypted features are stored in the database. The trained and test features are compared using Minutia Score matching algorithm. The proposed system provides False Matching Rate of 0.26 with False Non-Matching Rate of 0.45 without Secret Sharing Cryptographic Scheme. The finger print system with Secret Sharing Cryptographic Scheme offers False Matching Rate of 0.15 with False Non-Matching Rate of 0.30. The accuracy of the minutiae matching based finger print system provides 94.5%.

Keywords: Fingerprint, Pre-processing, Feature extraction, Matching, Databases