Enhancing Voting Security with Biometric Face Authentication: A Comprehensive System Design and Implementation

This study presents a biometric voting system that uses Convolutional Neural Networks (CNNs) for facial identification in order to improve electoral security and integrity. Election fairness is seriously threatened by traditional voting methods' susceptibility to problems like multiple voting, voter impersonation, and fake credentials. By using CNNs for accurate face matching and feature extraction, the suggested approach outperforms traditional techniques like Haar Cascade classifiers in terms of accuracy, resilience, and flexibility under a range of circumstances, such as changes in lighting, position, and occlusions. The system is also compared to other biometric methods as K-Nearest Neighbour (KNN), Principal Component Analysis (PCA), and Local Binary Patterns Histograms (LBPH). The CNN-based strategy obtains a greater recognition rate of 98.6%, according to experimental results utilizing a robust dataset, than these standard methods, which have an accuracy range of 80-90 %. To stop spoofing efforts and guarantee that only authorized users may take part, the system incorporates multi-factor authentication, including liveness detection techniques and One-Time Passwords (OTP). Large-scale election procedures are made possible by the architectural design's efficiency and scalability guarantees. The study also examines the ethical ramifications of biometric technologies in elections and data protection legislation compliance. The results of the experiment show notable gains in voter turnout, security, and user happiness, underscoring the system's potential as a safe substitute for traditional voting procedures. It might eventually be expanded to national or international elections with more research and testing.