Cancelable biometric template generation based on Log-Gabor filter and Hyperelliptic Curve Cryptography and Neuro fuzzy system

Abstract

For biometric identification systems to be secure and private, the creation of cancelable biometric templates has emerged as a critical field of study. This study suggests a new way to make biometric templates that can be deleted. It does this by using the strength of Log-Gabor filters for feature extraction, the safety of Hyperelliptic Curve Cryptography (HECC), and the accuracy of a neuro-fuzzy system for classification. Log-Gabor filters extract strong phase-based characteristics from iris pictures to achieve high discriminating capacity over a variety of spatial frequencies. We use a lightweight cryptographic approach, HECC, to effectively encrypt the retrieved characteristics and secure the biometric templates against unauthorized access and data breaches. Furthermore, we incorporate a neuro-fuzzy system (NFS) to enhance decision-making by combining the learning capabilities of neural networks with the reasoning flexibility of fuzzy logic. Tests on the benchmark CASIA Iris V4 and IITD iris datasets show that the suggested method achieves safe, reversible, and very accurate biometric authentication with a low false acceptance rate (FAR) and false rejection rate (FRR). In order to meet the demands of both current biometric systems. This hybrid technique provides a secure and efficient solution.