An AI-Enhanced Biometric Authentication Framework for Transparent Digital Elections

Abstract

The proposed SmartVote system is based on the idea of improving the security and robustness of the voting process using an artificial intelligence-based approach with dual biometric verification. The proposed SmartVote system is based on an optical fingerprint sensor and artificial intelligence-based facial recognition technology to verify the voter\'s ID. Only verified individuals can cast their votes using the proposed SmartVote system. Advanced features such as live detection and duplicate vote checking have been included in the proposed SmartVote system to avoid any type of spoofing, impersonation, or other types of fraudulent activities. The proposed SmartVote system is implemented using a Python programming language as the core language. Other libraries used in the proposed SmartVote system are OpenCV for image processing, MediaPipe for facial recognition, TensorFlow/Keras for machine learning algorithms, and PyQt5 for the GUI interface. Biometric verification is implemented using fingerprint matching algorithms and facial embedding algorithms. MySQL/SQLite is used for storing data securely in the proposed SmartVote system.

Introduction

The SmartVote system is an AI-based secure electronic voting platform designed to overcome issues in traditional e-voting such as fraud, impersonation, duplication, and weak identity verification. It uses dual biometric authentication (fingerprint and facial recognition) combined with artificial intelligence features like liveness detection, spoof prevention, and behavioral analysis to ensure that only genuine voters can cast votes.

The system includes voter registration, biometric data collection, preprocessing, AI model-based verification, and secure vote storage. It also adds monitoring, fraud detection, multilingual support, voice assistance, and time-slot-based voting to improve accessibility and manage voter flow. Votes are categorized using a risk-based system (green/yellow/red) for administrative monitoring, with real-time dashboards and logs for transparency.

The methodology involves collecting biometric data, extracting features, training AI models (CNN-based face recognition and fingerprint matching), and integrating all modules into a Python-based system using tools like OpenCV, TensorFlow/Keras, and PyQt5.

Literature shows that while biometric and AI-based voting systems exist, most lack strong security integration, explainability, or scalability. This project improves upon them by combining multiple security layers, AI-based verification, and fraud detection in a unified system.

The system aims to provide a secure, transparent, and efficient voting process, and future improvements include blockchain integration, cloud scalability, advanced biometrics, and mobile/remote voting capabilities.

Conclusion

The SmartVote system has been successfully designed and implemented, demonstrating efficient and reliable performance across all modules. The integration of dual biometric authentication using an optical fingerprint sensor and AI-based facial recognition ensures accurate voter verification and significantly reduces the risk of impersonation and duplicate voting. The inclusion of liveness detection further strengthens the system by preventing spoofing attempts and ensuring that only genuine users participate in the voting process. All components of the system, including voter enrollment, biometric processing, authentication, vote casting, and monitoring, are seamlessly integrated to provide a smooth and secure workflow. The system operates in real time with minimal delay, offering a user-friendly interface and efficient data management. The voting process is carried out securely, with all votes stored safely in the database while maintaining confidentiality and integrity. The results obtained from the implementation confirm that the system performs effectively and meets the intended objectives. The SmartVote system proves to be a robust, scalable, and practical solution for secure electronic voting. It can be successfully applied in institutional and small-scale elections, contributing to transparent, accurate, and trustworthy election management.

References

[1] Maltoni, D., Maio, D., Jain, A. K., and Prabhakar, S. (2009). Handbook of Fingerprint Recognition. Springer. A foundational reference on fingerprint extraction, minutiae matching, and biometric security. [2] Zhang, K., Zhang, Z., Li, Z., and Qiao, Y. (2016). “Joint Face Detection and Alignment Using Multi-task Cascaded Convolutional Networks (MTCNN).” IEEE Signal Processing Letters, 23(10), 1499–1503. A widely used model for fast, accurate face detection. [3] Parkhi, O. M., Vedaldi, A., and Zisserman, A. (2015). “Deep Face Recognition.” British Machine Vision Conference(BMVC).Introduces the VGGFace CNN model used in high-precision face recognition systems. [4] Nikisins, O., and Petrovics, G. (2018). “Presentation Attack Detection in Biometric Systems: A Review.” IEEE Access,6,13530–13549.Comprehensive review on liveness detection for anti-spoofing in face and fingerprint recognition. [5] McCorry, P., Shahandashti, S. F., and Hao, F. (2017). “A Smart Contract for Boardroom Voting with Maximum Voter Privacy.” International Conference on Financial Cryptography and Data Security. Discusses blockchain and cryptographic protocols relevant for secure vote storage and auditing. [6] Stallings, W. (2017). Cryptography and Network Security: Principles and Practice. Pearson. Key reference for AES, RSA, hashing, message authentication codes, and secure communication. [7] Ross, A., Nandakumar, K., and Jain, A. K. (2006). Handbook of Multibiometrics. Springer. Covers multimodal authentication systems such as combined fingerprint and face verification. [8] Chollet, F. (2017). “Xception: Deep Learning with Depthwise Separable Convolutions.” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1251–1258.CNN architecture used in high-performance face recognition and liveness detection models. [9] Adida, B. (2008). “Helios: Web-based Open-Audit Voting System.” USENIX Security Symposium, 335–348. A transparent and verifiable e-voting model, useful for understanding secure online election frameworks. [10] Olaniyi, O. M., Folorunso, T. A., Ahmed, A., and Joseph, O. (2016). “Design of Secure Electronic Voting System Using Fingerprint Biometrics and Crypto-Watermarking Approach.” International Journal of Computer Applications. Discusses fingerprint-based voting and encrypted watermarking for vote integrity.of Environmental Research and Public Health, 19(19), 12633.

Copyright

Copyright © 2026 Mrs. J. Sowmya, K. Sai Krishna, D. Santhosh, M. Varshith Yadav . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.