Real-time Face Recognition System with Email for Enhanced Security

Abstract

This project presents an automated face classification system designed to identify individuals from input images and send the matched results directly to a specified email address. The workflow integrates image preprocessing, feature extraction, embedding comparison, and automated communication into a single pipeline. A simple Flask interface manages user registration, updates, and deletion of stored records. Experimental results show strong classification accuracy and reliable performance across varied input conditions. The proposed solution offers an efficient and deployable framework suitable for real-time identification in attendance systems, small-scale surveillance, and operational monitoring.

Introduction

The text presents an automated facial recognition system designed to improve practicality, efficiency, and integration in real-world security and verification applications. While existing facial recognition methods achieve high accuracy, many rely on manual processes or fragmented workflows that separate recognition, reporting, and result delivery. This project addresses these limitations by developing a streamlined system that not only classifies faces using embedding-based recognition but also automatically sends matched images to the associated email address and displays results on a centralized dashboard.

Building on prior research, the system adopts modern deep learning–based facial embeddings rather than traditional statistical approaches. Lightweight and efficient embedding models are used to balance accuracy with low computational cost, making the solution suitable for small-scale and real-time environments. Unlike earlier systems that focus only on recognition accuracy, this approach emphasizes automation, reporting, and usability by integrating recognition, logging, notification, and visualization into a single workflow.

The methodology involves image acquisition and preprocessing, face embedding generation, similarity-based classification, database logging, automated email notification, and dashboard visualization. All processes are triggered automatically when images are placed in an input folder, eliminating manual intervention. An SQLite database stores classification results, while a Flask-based dashboard provides a simple interface for viewing outcomes.

The system architecture is modular, consisting of a Flask application for user management, a standalone classification and email module, structured image storage, and a database-driven dashboard. Performance evaluation shows strong results, achieving 94.7% accuracy with high precision, recall, and efficient processing time. Overall, the project demonstrates a practical, integrated facial recognition solution suitable for applications such as attendance tracking, access control, and small-scale security monitoring.

Conclusion

The developed face classification system provides a complete and functional pipeline that integrates automated identification with email-based delivery of matched results. By combining preprocessing, reliable feature extraction, and accurate embedding comparison, the system achieves consistent performance across various input conditions while reducing the need for manual effort. Its unified workflow, which merges recognition and communication into a single process, offers a practical benefit compared to conventional methods that treat these steps separately. Although the system might encounter difficulties when dealing with heavily occluded, poorly illuminated, or low-resolution images, it remains dependable for most realworld scenarios. Future enhancements such as advanced augmentation techniques or adaptive thresholds can further improve its robustness. Overall, the system presents an efficient and scalable solution suitable for applications in monitoring, attendance management, and security operations.

References

[1] “Build Your Own Face Recognition Service Using Amazon Rekognition,” AWS Blog, 14 Aug. 2017. [2] L. S. Bhattad, K. B. Bijwe, and V. B. Bhagat, “Review of an Efficient Face Recognition System Using Hybrid Methodology,” Master of Engineering, vol. 5, no. 1, 2017. [3] A. P. Singh, “Image Spam Classification using Deep Learning,” 2018. [4] A. P. Singh, S. S. Manvi, P. Nimbal, and G. K. Shyam, “Face Recognition System Based on LBPH Algorithm,” International Journal, vol. 8, no. 5S, May 2019. [5] F. Deeba, A. Ahmed, H. Memon, F. A. Dharejo, and A. Ghaffar, “LBPH-based Enhanced Real-Time Face Recognition,” vol. 10, no. 5, 2019. [6] Babuakash, “Face Recognition Automation and Sending Emails and WhatsApp Messages via Face Detection,” Article, 24 June 2021. [7] R. Kumar, “An Overview on Amazon Rekognition Technology,” 2021. [8] B. Balabharathi and Y. Yuvasri, “Face Recognition System – LBPH Algorithm,” May 2022. [9] P. Gowsikraja, T. Thevakumaresh, M. Raveena, J. Santhiya, and A. R. R. Vaishali, “Object Detection Using Haar Cascade Machine Learning Algorithm,” International Journal, vol. 10, no. 6, June 2022. [10] M. Kaur and N. Dhillon, “Face Recognition using LBPH Algorithm, Python and OpenCV,” vol. 4, no. 9, Sep. 2022. [11] A. Basitty, “A Review Paper On Face Recognition Using LBPH Technique,” vol. 5, no. 5, May 2023. [12] P. Futane, P. Shelke, A. Khedkar, T. Joshi, S. Chaudhari, and C. Shewale, “A Real-Time Face Recognition System with Email and WhatsApp Integration for Enhanced Security,” 2023. [13] R. Tandon, A. Sayed, and M. A. Hashmi, “Face Mask Detection Model Based on Deep CNN Technique Using AWS,” 2023. [14] J. B. Urbanus and Y. M. Malgwi, “Software Framework for Face Recognition System Using Amazon Rekognition,” vol. 10, no. 5, pp. 135–154, 2024. [15] N. El Fadel, “Facial Recognition Algorithms: A Systematic Literature Review,” 2025.

Copyright

Copyright © 2025 Mrs. Radhika S N, Anvitha H M, Chethana T, Impana T S, Apeksha K K. 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.