FKP and Iris Based Multimodal Biometric System Using PCA with NFNN

Abstract

The Biometric system is an important pattern-based identification system that is widely incorporated in various sectors for authenticating human beings using various biometric traits. Usage of any biometric authentication system with only single biometric trait may not be accurate enough to provide desired results with maximum sensitivity and productivity. Several biometric features, such as fingerprints, palm veins recognition, palm, hand geometry, iris recognition, DNA used to authenticate the user's identity. From the various biometric features, the finger-knuckle print (FKP) and Iris have the fine, rich texture. There are also stable features in the FKP and iris, which can rarely be broken by the intermediate person. So, the proposed system uses the FKP and Iris as the biometric feature while analyzing the authentication in the various applications. The texture pattern present in Finger Knuckle print and Iris are distinct when combining both the patterns will become highly unique. FKP and Iris images have pre-processed by using the Gabor filter and the exact regions are segmented using the edge with region of interest method. From the extracted region different features are extracted with the help of principal component analysis. Both the extracted features were fused at score level. Finally the matching is performed with the help of the Neuro fuzzy neural network (NFNN). The performance was evaluated with dual properties extracted from PolyU FKP database and CASIA Iris database. The usefulness of the proposed design is measured in terms of False Acceptance Rate (FAR), False Rejection Rate (FRR), Equal Error Rate (EER) and accuracy.