Face Recognition Security Camera and Intruder Detector System based on Internet Of Things

A college project about the implementation of facial recognition and the internet of things to a security camera system. A security camera system that detects an unidentified face and sending an e-mail as an alert.

By Niko Wahyudi

Prerequisites

Windows 10 or Linux
Python 3
Python Packages Manager (PIP)

Getting Started

Installation

Clone this repository:

git clone https://github.com/wahyudiniko/security-camera-intruder-detector.git
cd security-camera-intruder-detector

Installing Requirement Packages

pip install -r requirements.txt

For windows 10, can install OpenCV via pip

pip install opencv-python
pip install opencv-contrib-python

For Raspbian/Raspberry OS or other Linux distros you can follow the installaion guide here or run the installation_script.sh bash script to install some dependencies.

Usage

Registering Face

Run the file face_shot.py in the terminal:

python face_shot.py

Insert your name

and then a window will pop up.

Press spacebar to take the picture or press Esc to abort the program.

Modify Sender and Reciever E-mail

After the requirements were installed, you can modify the to variable to the e-mail address you want to use to recieve the e-mail and the UserName variable for the sender e-mail. For UserPassword variable, you can generate an app password here to avoid storing your password in plain text. here's the code in the mail.py file line 9-11

to = "receiver_email_address"           # receiver email address
UserName = "sender_email_address"       # sender e-mail address
UserPassword = "sender_email_password"  # both sender and receiver e-mail address only works for gmail

Modify Login Authorization

In order to secure the webcam access, we need authorization.

Modify the code to make your own authorization account in the main.py file line 13-14:

app.config['BASIC_AUTH_USERNAME'] = 'your_uname'
app.config['BASIC_AUTH_PASSWORD'] = 'your_password'

Execution

You can run the main.py program by typing this command in the command line or terminal:

python main.py

Wait until this displayed on command line

Accessing Webcam From Web-browser

Type https://your_local_ip_address:5000 inside the address bar. Then an authorization form pops up. Fill it with your authorization username and password.

This is the web-browser display in PC.

When faces are detected the frame rate will significantly drop and it's slower than when not detecting faces.

You can also access it from different devices that connected to the same network. You also being asked for a login authorization. This is the web-browser display in smartphone.

This is the log of devices that are currently accessing the webcam through a web browser.

Alert

When an unidentified person is detected, an e-mail will be sending to the receiver email.

This is the display in web browser.

This is a status update being displayed in the command line.

An alert e-mail has been sended.