Digital Voting System

Abstract

The Secure Biometric-Based Online Voting System proposed in this project addresses the main limitations of Pakistan’s traditional paper-based electoral process, including identity fraud, ballot manipulation, manual counting delays, and restricted voter acces sibility. The system implements multi-factor authentication through email OTP verifi cation, live face recognition using ArcFace and OpenCV, and camera-based fingerprint verification to ensure accurate voter identification and enforce a strict one-person-one vote policy. CNIC-based automated constituency detection enables users to cast NA and MPA votes online through a Flutter-based interface, supported by a Flask backend de ployed on Azure and a SQL Server database. Votes are recorded in a blockchain ledger to guaranty immutability and auditability, while an admin dashboard enables monitoring of flagged biometric mismatches and suspicious registrations. Extensive testing—including performance, usability, security, GUI, and exception handling—demonstrated high reli ability, strong security, and smooth user experience. In general, the system provides a scalable, transparent and tamper-resistant digital voting framework suitable for future nationwide adoption, with potential enhancements such as NADRA integration, advanced biometrics, and offline polling support.