A Study on Adaptive Biometric Fusion Strategy and its Application on Biometric Template Protection

This research explores a biometric authentication framework utilizing hashing techniques like Alignment-Free Hashing (AFH) and Index-of-Max (IoM) hashing, and feature-level fusion for template protection. It ensures robust security, accommodates diverse biometric features, and achieves high performance. Additionally, a novel key-based fusion method enhances security through the introduction of fusion key and encryption. A multimodal biometric cryptosystem, integrating AFH, IoM hashing, and the key-based fusion method, is developed, supporting various template sizes and representations. Extensive testing validates the effectiveness of the proposed schemes.