Analysis of decision level fusion in multimodal biometrics using iris and fingerprint

Biometrics is the most advance technology for identifying any person. It is an authentication technique which place confidence in measurable individual and physiological characteristics that will be mechanically verified. These systems can operate either in identification or verification mode. Because security breaches and dealings fraud have increased much in level, the necessity of technologies for extremely secure identification and private verification is changing into apparent. Due to some limitations of unimodal biometric system, multimodal biometrics has been introduced where fusion of the modalities is the bigger challenge. In Multimodal Biometrics, Fusion can be performed on different levels. This research evaluates the performance of multimodal biometrics using three different fusion approaches with logical AND and OR operator: K-Nearest Neighbour, Hidden Markov Model, and Neural based Classification. The recognition of the iris and fingerprint based multimodal has been fused using decision level fusion. Various experiments results into decision level fusion which is based on neural is best to be followed by AND. The system performance is evaluated in terms of Recognition Accuracy, False Acceptance Rate (FAR) and False Rejection Rate (FRR).