Quantum Cryptography with Optical Encryption for Cancelable Biometric Systems in Crowd Control

Crowd control is one of the new research trends in images and video processing. Crowds occur in cases in which so many people are allowed to enter a limited area as in Holy places, stadiums, events and demonstrations. There are two trends to deal with issue of crowds. The first trend is to use crowd analysis algorithms to analyze images or videos. The objective of this trend is to send alarms for operators or organizers to control the entry rate to crowded places. On the other hand, the second trend is to use biometrics from the beginning to limit the entry to allowed persons only. In case of using biometrics, the ethical dilemma arises of collecting user biometrics in large databases that may subject to hacking attempts. This paper presents a new trend for biometrics based on cancelable biometrics concepts to control the entry to crowded places. The objective is to use encrypted or deformed versions of people’s biometrics for the access process to avoid hacking attempts. The proposed approach depends on quantum image cryptography. The users’ face images are encrypted with quantum cryptography algorithms and saved in biometric databases. The objective of this process is twofold. First, the original templates are encrypted for privacy presentation. Second, quantum processing abilities are well-exploited to generate the templates efficiently. A correlation-based approach is exploited for template comparison and identity verification. Simulation results on FERET database, based on genuine and imposter distribution, ensure Equal Error Rate ( EER ) values of 100%, and Area Under the Receiver Operating Characteristic Curve (AROC) values of 99.51%. Moreover, experiments reveal high robustness to noise in addition to superior performance compared to other state-of-the-art algorithms.