Face Recognition Attendance System
IoT ESP32-CAM Wireless Streaming & OpenCV Recognition System
Wireless ESP32-CAM video capture, real-time face matching, automated Excel logging, and email intrusion alerts.
System Specs
Compiled Stack
Project Overview
The Face Recognition Attendance System is an integrated IoT security and management platform. An ESP32-CAM module, programmed in Arduino C, captures high-resolution video frames and streams them wirelessly over local IP addresses. A backend Python server captures this live stream, processes each frame via OpenCV to identify and match faces against a local database, logs verified attendance dynamically into Excel sheets, and dispatches automated email alerts (with attached stranger face snapshots) to administrators upon detecting an unknown person.
Application Architecture & Modules
ESP32-CAM Live Stream
Programmed the ESP32-CAM in Arduino C to initialize a Wi-Fi server and stream MJPEG video feeds over a dedicated local IP address with optimized frame rates.
OpenCV Face Detection
Configured real-time frame extraction and face detection algorithms utilizing OpenCV and deep learning face recognition models to match faces with extremely high confidence.
Excel Database Logger
Integrated python-openpyxl routines to construct, log, and maintain daily timestamped attendance logs automatically without requiring complex external SQL overhead.
SMTP Intrusion Alerts
Built a Python-based SMTP automation module that triggers instantaneous Gmail alerts to administrators when unrecognized faces are detected, complete with live snapshot attachments of the stranger.
Daily PDF / Excel Reports
Generates consolidated daily attendance reports summarizing active users, entry timings, and security logs, formatted for direct administrative export.
Low-Power Optimization
Configured sleep states on the ESP32 chip to wake up only when motion or PIR sensors detect changes, extending hardware battery longevity.
The Problem It Solves
Traditional attendance tracking (manual registers or card tapping) is prone to buddy punching, data errors, and slow aggregation. Furthermore, standard security systems only record footage without actively alerting admins to strangers. This project solves both by coupling wireless ESP32-CAM nodes with local OpenCV logic to instantly log verified faces to Excel and dispatch stranger alert emails with attached snapshots.
Real-World Uses
- Contactless, automated attendance tracking in offices, classrooms, or secure labs.
- Instant intruder detection system with active SMTP notifications and stranger face capture.
- Low-overhead standalone security monitoring for small businesses without expensive cloud subscriptions.
Want to build this or acquire the source?
Whether you are looking for the full production source code, circuit schematic diagrams, component list guidance, or custom feature integration, feel free to reach out. We are happy to help you build or customize this project.