Dual-Factor Smart Lock System Using Arduino Passcode and Raspberry Pi Facial Recognition

This repo implements a dual-factor smart lock: an Arduino handles a local passcode keypad and LEDs; a Raspberry Pi (or any Linux host) performs facial recognition with a camera and acts as the second factor. The Arduino and Pi talk over USB serial: the Arduino requests recognition, the Pi runs the recognition and replies with an unlock/deny message.

The README below is intentionally concise and focused on what you need to build and run the system on a Raspberry Pi.

Quick summary

• Arduino: reads 5 passcode buttons, Enter button, and a Send/Recognition button. Controls Red/Green/Blue LEDs. Sketch: ArduinoFaceLock/ArduinoFaceLock.ino.
• Raspberry Pi: runs the Python host main.py, captures camera frames and performs recognition using face_recognition/dlib. Replies to Arduino with FACE_RECOGNIZED or FACE_UNRECOGNIZED.
• Face data: persisted to faces.json (name -> encoding list).

Circuit (ASCII map)

Below is a simple ASCII diagram showing how to wire components to an Arduino (Uno/Nano). All buttons use INPUT_PULLUP, so the other side of each button connects to GND. LEDs use series resistors (about 220Ω) to GND.

                 +5V
                  |
                 [ ]  (not used directly; buttons use INPUT_PULLUP)

Arduino UNO / Nano
-------------------
            _______
           |       |
  D2 ----o-| BTN1  |
  D3 ----o-| BTN2  |
  D4 ----o-| BTN3  |
  D5 ----o-| BTN4  |
  D6 ----o-| BTN5  |
           |       |
  D7 ----o-| ENTER |
  D8 ----o-| SEND  |---(USB)----> Raspberry Pi (serial) ---(USB)----> CAMERA
           |       |
  D9 ----[220Ω]----|> RED LED  -> GND
 D10 ----[220Ω]----|> GREEN LED-> GND
 D11 ----[220Ω]----|> BLUE LED -> GND
           |_______|

Buttons: other side -> GND
LEDs: cathode -> GND, anode -> resistor -> digital pin

USB: connect the Arduino to the Raspberry Pi's USB port. The Pi will see a
serial device (commonly /dev/ttyACM0 or /dev/ttyUSB0).

Raspberry Pi host: quick setup

1. Hardware: connect a Pi camera or USB webcam and the Arduino over USB.
2. OS: Raspberry Pi OS (or similar). Make sure the camera works (e.g. with libcamera or OpenCV).
3. Python: create a virtualenv and install dependencies:

python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install face_recognition numpy opencv-python pyserial

4. Run the host program (adjust port if necessary):

python3 main.py --port /dev/ttyACM0 --baud 9600 --timeout 10

On the Pi, the serial device is commonly /dev/ttyACM0 (Uno/Leonardo) or /dev/ttyUSB0 (some USB-serial adapters). If you get a permission error, add your user to the dialout group: sudo usermod -aG dialout $USER and re-login.

Enrolling faces (simplified)

Use this short Python snippet on the Raspberry Pi to add a face from an image file. The repo contains faces.FaceManager which handles persistence.

from faces import FaceManager
import face_recognition

fm = FaceManager('faces.json')
img = face_recognition.load_image_file('alice.jpg')
enc = face_recognition.face_encodings(img)
if enc:
    fm.enroll('alice', enc[0])
    print('enrolled alice')
else:
    print('no face found')

After this, the Pi will recognize "alice" when the camera sees her face and the Arduino has requested recognition.

How it works (runtime flow)

1. User either enters a correct passcode locally (Arduino only) or presses the SEND/Recognition button to request facial recognition.
2. If SEND is pressed, Arduino prints RECOGNITION_MODE on serial and turns the Blue LED on.
3. Raspberry Pi receives the request, captures camera frames and tries to match against enrolled faces using stored encodings. If a match is found within the timeout, Pi sends FACE_RECOGNIZED, otherwise FACE_UNRECOGNIZED.
4. Arduino receives the response and either unlocks (Green LED) or shows an error (Red LED).