A modern, secure, and fully automated door-locking system powered by facial recognition. This project combines embedded systems, computer vision, and IoT principles to deliver a reliable access-control mechanism designed for real-world deployment. It’s the kind of solution that politely tells traditional keys, “Your services are no longer required.”
This system uses a camera module and a face recognition algorithm to identify authorized users and unlock the door automatically. Unauthorized attempts trigger a security response.
The whole architecture is engineered to be: Accurate Automated Tamper-resistant User-friendly Scalable for home, office, or lab environments
Real-time Face Detection & Recognition Identifies pre-registered users with high accuracy.
Automatic Door Unlocking Uses a relay/servo/motor-based locking mechanism.
Security Alerts Unauthorized users trigger buzzer alerts or warning messages.
Local + Cloud Storage (optional) Store face datasets locally or sync with cloud.
Low-power Embedded Hardware Designed for efficient performance in embedded environments.
Layer Technology Hardware ESP32 / Raspberry Pi / Camera Module, Relay/Servo Motor Software Python / OpenCV OR ESP-IDF/Arduino (depending on implementation) Recognition LBPH / Haar Cascade / CNN-based model Storage Local File System / SD Card / Cloud Integration
Smart-Face-Recognition-Door-Lock/ │── /dataset → Stored images for training │── /training → Model training scripts │── /src → Main firmware/software code │── /hardware → Schematics, connections, parts list │── /docs → Documentation and reports │── README.md → You’re reading it :)
Camera scans the face of the person standing at the door. The system extracts facial features and matches them with the trained model. If the face is recognized → Lock opens automatically. If the face is not recognized → System triggers alert mechanisms. The entire flow runs autonomously with minimal user intervention. Classic workflow, future-proof execution.
ESP32-CAM / Raspberry Pi (depending on design) OV2640 Camera Module Relay Module / Servo Motor Lock Buzzer (for alerts) Power Supply (5V/2A) Jumper Wires & Basic Hardware Tools
Python 3.x (if Raspberry Pi version) OpenCV Arduino IDE / ESP-IDF (for ESP32 version) Serial Monitor Tools Training Scripts (for dataset generation)
-
Prepare the Hardware Connect camera module to ESP32/RPi Connect relay/servo to control the lock Power up and verify connections
-
Dataset Creation Run: python capture_images.py
Capture 50–100 images per authorized user.
-
Train the Model python train_model.py
-
Deploy Firmware Flash the microcontroller or run the Python script on Raspberry Pi.
-
Test the System Stand in front of the camera → observe door lock response. Straightforward. Mission-critical. No drama.
Mobile App to manage user access Cloud-based face dataset Two-factor authentication (Face + OTP) Integration with smart home systems Night-vision support
Ayush Gupta 20233096(Group Leader) Ayush Joshi 20233097 Ayush Pandey 20233331 Chandrakant Suman 20233106 Gopal Maheshwari 20233336 Harsh Agrawal 20233522 Harsh Chaurasiya 20233136 Jitendra Kumar 20233150 Kanishk Agrawal 20233526 Keshav Kumar 20233339 Keshav Kumar 20233158
Embedded Systems Project Subject Code Number: CSN15402
Course Coordinator: Saugata Roy