An Approach for Multimodal Biometric Authentication using Genetic Algorithm
Amber Hayat, Kumar Abhishek, Ashok KumarBhateja, Saibal Kumar Pal · 2024
Multimodal biometric authentication is a robust security mechanism designed to enhance user authentication systems’ reliability and security. It includes two or more biometric traits to design a more reliable and secure identification system. The enhanced accuracy and robust security make multimodal biometric authentication an excellent choice for improving the security and performance of user authentication systems. This paper proposes a multimodal authentication approach fusing fingerprint, ear, and face biometrics. Further, a dynamic weightage is assigned to each of the matching systems, calculated with experimental analysis using a Genetic algorithm. The genetic algorithm has shown excellent results in the proposed approach owing to its ability to work with noisy data and explores the search space to give effective results. The publicly available face data ORL, Yale, IITD Ear data, and custom datasets are utilized. The approach is evaluated on the above datasets and achieved an accuracy of $97 \%$.