Towards Electronic Voting Using Face Verification Technology

This study presents an electronic voting system using the face detection and verification methods. The proposed system consists of face detection and face identification module to detect and authenticate voters' face. Initially, voter's face is detected using facial land mark estimation module. While face verification module uses linear discriminant analysis integrated with the adaptive boosting method. Face detection and verification modules work perfectly up to pose variation up to ±15°. Detailed experiments show that proposed system achieves a mean accuracy of 94.15% and requires nearly 50 seconds to cast a single vote. The study also contributes to the development of an excellent alternative to conventional electoral system.