https://www.irjmets.com/upload_newfiles/irjmets70800025919/paper_file/irjmets70800025919.pdf

India's election process continues to face key issues such as voter impersonation, repeated voting, and insufficient transparency.To address these problems, this project introduces an AI-powered system for Election Fraud Detection and Post-Election Voter Analytics.The system captures live facial images of voters through webcams at polling stations.Using deep learning models specifically face-api.jswith the faceRecognitionNet the system extracts distinct facial features and generates encrypted face embeddings.These are stored in a database and compared in real-time through a 1: N matching process using Euclidean distance, ensuring no duplicate registrations or cross-booth voting occur.Testing with simulated data confirmed successful implementation of the facial recognition module.If a matching face is found, registration is blocked; otherwise, the system proceeds with new voter registration.In addition, the platform supports automatic validation of biometric-verified voter counts against EVM data, offers a post-election analytics dashboard, and provides a public-facing portal for voter verification and grievance reporting.Overall, the system presents a scalable, secure, and transparent approach to modernizing the electoral process.