Key points are not available for this paper at this time.
Fingerprint and face recognition are one of the most widely used biometric approaches. Detection of fingerprint has become one of the most popular ways of maintaining security system in today's technological era. Most Authentication algorithm on the internet and research papers suffer from spoofing attacks. This paper proposes a robust authentication system that integrates two biometric technologies, fingerprint and face recognition, for enhanced security. The proposed system employs a multi-level authentication approach, where the user's identity is verified through face recognition and fingerprint authentication. Additionally, the system utilizes facial emotion detection to enhance the security of the system by ensuring that the user's emotional state is consistent with their identity. The process initiates by acquiring the facial image of the user and cross-referencing it with the stored facial template in the database. If the face match is successful, the system will proceed to verify the user's fingerprint. Upon a successful fingerprint match, the system will employ facial emotion detection to assess the user's emotional state. If the system deems the user's emotional state suitable, it will then prompt the user to respond to a security question. The security question is randomly generated and can be customized based on the system's requirements. If the user answers the question correctly, the system will grant access to the real database. However, if the answer is incorrect, the system will redirect the user to a dummy dataset to prevent unauthorized access. The proposed system offers several advantages over traditional authentication systems, including increased security, accuracy, and convenience. By integrating multiple biometric technologies, the system offers robust protection against identity theft and unauthorized access. Additionally, the use of facial emotion detection and security questions enhances the security of the system by providing additional layers of authentication of 97.29%
Shanmugapriya et al. (Thu,) studied this question.