Biometric technologies are now a cornerstone in the evolving landscape of security systems. However, persistent threats such as hacking and the potential exposure of biometric data during cyberattacks highlight the pressing need for secure and innovative authentication mechanisms. One promising direction is the use of cancelable biometrics, which offer an added layer of protection for biometric templates.
In this study, we introduce a novel variant of cancelable biometrics that leverages the robust capabilities of the Double Random Phase Encoding (DRPE) algorithm based on Quaternion Mathematics. By integrating an additional biometric input and utilizing quaternion-based computations, our method achieves outstanding results that go beyond existing benchmarks.
The cancelable face recognition framework we propose maintains impressive reliability under noisy conditions, as confirmed by simulation tests. Specifically, our system achieves an area under the ROC curve of 0.994 and an exceptionally low Equal Error Rate (EER) of just 0.00174. Furthermore, it exhibits remarkable efficiency, generating each secure template in approximately 0.056263 seconds. These results underscore the solution’s high performance and suitability for tackling modern security challenges.
Autors: Mahmoud Nasr, Adam Piórkowski, Fathi E. Abd El-Samie
DOI: 10.1007/s11042-024-18621-1
Keywords: Quaternion Mathematics
READ HERE