PPG Based Biometric Authentication System

Recent advancements in biosensing, particularly in photoplethysmography (PPG), have spurred the adoption of biometric authentication, supplanting traditional password-based systems. PPG, renowned for its non-invasive monitoring of blood flow changes, notably via wearable devices, presents a viable avenue for continuous user verification. This paper introduces an innovative authentication system harnessing PPG technology, ensuring ongoing validation compared to conventional methods. The system entails preprocessing stages encompassing noise removal, quality screening, and feature extraction, bifurcated into Training and Testing phases. During Training, processes such as noise removal, feature extraction, and encryption (Homomorphic) bolster system accuracy and security. The Testing phase mirrors these procedures, followed by secure data storage and analysis using a Homomorphic classifier. Through this framework, the system not only enhances security but also usability in biometric authentication. By offering continuous verification and leveraging PPG, it ensures a seamless user experience while fortifying protection against unauthorized access. The proposed system provides an accuracy of 98% with EER of 0.25 with PUT Vein dataset.