System evaluation demonstrates high authentication accuracy and decentralized record management in electronic voting systems, indicating improved resistance to electoral fraud.
Key Points
To develop and evaluate a decentralized, tamper-resistant voter accreditation framework that integrates multimodal biometric verification with private blockchain architecture on low-cost hardware.
Integrated dual-biometric authentication (facial and fingerprint recognition) with a private Ethereum blockchain network and standard cryptographic mechanisms.
Implemented and evaluated the architecture on a Raspberry Pi 5 embedded processing platform to test feasibility on low-cost hardware.
Fingerprint recognition demonstrated an average authentication accuracy of 97.8%, and facial recognition achieved an average accuracy of 95.1%.
System reliability evaluation demonstrated operational uptime exceeding 95% across the testing duration.
Blockchain storage overhead exhibited near-linear growth relative to transaction volume, maintaining stable and predictable transaction processing.