Facial-recognition-based-automatic-door-lock-unlock-system

Introduction

This project aims at automating the locking and unlocking of the main door of the house. Camera installed at the main door is used to click 5 consecutive photos of the person standing in front of the door as he/she presses a button present on the main door. The images thus captured, is processed by the Python Facial Recognition software, which then detects any known faces in the picture and unlocks the door for the people whose faces are identified successfully. However, if there are no known faces in the picture taken, then the picture is forwarded to the owner through Telegram Bot. The owner then may choose to let the unknown guests in through a simple telegram command, which sends a high signal to the arduino which opens the door. If the owner chooses to keep the door closed, then he does not reply to the bot which keeps the door locked. If the owner himself is standing in front of the door, his image is identified by face_recognition library, which signals the arduino to open the door. Also the owner can decide whether the identified person has the access authority to take other people inside with him or not.

Dependencies:

1. Python 3
2. dlib (Required to install face_recognition)
3. face_recognition
4. Telegram API
5. BeautifulSoup

Hardware Dependencies

1. Arduino
2. L293D IC
3. DC Motor
4. Connecting Wires
5. DC Power Supply

Installation for Python 3

Visit: https://www.python.org/downloads/ and download the version suitable for your System

Installation for face_recognition

This module is tough to install as installing dlib, which is a dependency for this module is quite challenging. Linux users may have some easy in installation as they can use pip to install them directly. But windows users are recommended to go through the following tutorial for the installation process.

ageitgey/face_recognition#175 (comment)

Other useful links are listed below:

1. https://pypi.org/project/dlib/
2. https://pypi.org/project/face_recognition/

Installation for Telegram API

Open Command Prompt (Windows) or Terminal (Linux) and type in the following command:

 pip install python-telegram-bot

Visit: https://python-telegram-bot.org/

Installation for BeautifulSoup

Open Command Prompt (Windows) or Terminal (Linux) and type in the following command:

 pip install bs4

Visit: https://www.crummy.com/software/BeautifulSoup/

Interfacing Arduino

The arduino used here is Arduino Uno and is programmed with Arduino IDE. Further information can be found in the following link. https://www.arduino.cc/

L293D IC

This is a motor driver IC which is primarily used to pass high-current and high-voltage to the motor but control the inputs depending on the signals from the Arduino. Since the Arduino is restricted to supplying of 5V and cannot meet the current requirement to drive a DC motor, L293D IC has been used. The IC Datasheet can be found at http://www.ti.com/lit/ds/symlink/l293.pdf

DC Power Supply

For this, any 5-12 volts power supply can be used. For the purpose of this project, we have used Smartphone Charger.

How to use?

1. Program the Arduino with the given code
2. Connect the wires and set up as shown in the circuit diagram
3. Create an account in way2sms (http://www.way2sms.com/)
4. Create a Telegram bot using BotFather
5. Add the missing credentials and Token in the python files (face_recg.py and TelegramAPI.py)
6. Add a well lit picture of the person whose face has to be recognized in the Assets folder and change face_recg.py accordingly
7. Run the TelegramAPI.py and keep it running
8. Run the face_recg.py when ready for facial recognition
9. Interact with the bot with the given commands
10. To get started, you can send /help to the bot