An Enhanced ATM Security System Using Multimodal Biometric Strategy

This paper presents a new biometric identification and authentication schema in relation to payment systems and ATMs. The financial sector has used ATMs as a means to make payment and offer financial services for its clients. But, security is a major issue in accessing these machines. Improving the performance of individual matchers in the aforementioned situation may not be effective. Multi-biometric systems are used to overcome this problem by providing multiple pieces of evidence of the same identity. In this paper, we have proposed the development of a fingerprint and iris fusion system which utilizes a Minutiae Matcher for fingerprint and Hamming Distance Matcher for iris with matching score level. It has been found that the proposed multimodal technique using threshold of 0.6 gave the best results. It has an accuracy of 96.67%, FAR of 0%, FRR of 5%, and run time of 32 seconds. Index Term-Multi-model System, Biometric, Fingerprint, Iris, Identification, Recognition, ATM, Biometric Computing, Fingerprint Recognition, IRIS Recognition, Minutiae Extraction, FAR, FRR.