Facial Recognition and Login Facial System

Stephany Valderrama and Wenya Li  

Thanks to: Aaron Moran , Arnas Steponavicius and Thomas Kenny. Modified code from those repositories

Requirements

The requirements needed for this project are as follows :

• Python 3.8+

• Linux, Windows, MacOS

• Django

• Latest version of pip

Installation

For run theses facial recognition demos, you require to install the following python module.

You can run pip install requirements.txt

pip3 install django
pip3 install numpy
pip3 install opencv-python
pip3 install Pillow
pip3 install face_recognition

Note: Make sure you are using the latest version of pip.

• For Mac or Linux --> follow the instructions of face_recognition docs

• For Windows --> follow the instructions of face_recognition docs

Note: Currently the packages for face_recognition are not fully supported on Windows.

Estructura de este repositorio

Carpeta	Folder / Fichero	Descripción
facial_recognition_webcam	facial.py	Face recognition system on live video from a local webcam.
facial_recognition_webcam	/info_facial	Imágenes de referencia para reconocimiento facial múltiple.
login_recognition_app	/accounts	Codigo html para el registro de usuarios.
login_recognition_app	/facialrecognition	Configuracion de metodos, funciones y conexion entre modulos par despliegue en Django.
login_recognition_app	~/ /media	Aquí se almacenan las imágenes de los usuarios registrados.
login_recognition_app	db.sqlite3	base de datos SQlite donde se guardan las migraciones de los nuevos usuarios registrados.
login_recognition_app	manage.py	Archivo principal para el funcionamiento del sistema de login.
	.gitignore	Para ignorar los archivos sensibles como imégenes.
	requirements.txt	Librerías y paquetes utilizados par la creación de este proyecto.
	README.md	Descripción de este repositorio

Demos with face recognition and openCV libraries.

We proposed two demos using face recognition library:

• Simple sample of Face recognition system on live video from a local webcam.

• Facial login web application with Django framework and sqlite databases.

Running the Face recognition system on live video from a local webcam

This is a face recognition system on live video from a local webcam.

1. Navigate to directory

   $ cd \facial_recognition_webcam\
   

2. Put into \info_facial folder the images of the persons that you want to recognize in this demo.

The image format could be .jpg, .jpeg, .png.

4. Run the application:

   $ python facial.py 
   

5. Look at the camera for take the picture to recognize faces and compare with the base images of the system.

6. Press CTRL-C to stop the process.

Running the Facial login web application Locally

1. Navigate to directory

   $ cd \login_recognition_app\facialrecognition
   

2. Make migrations of the required models and tables needed to run the program.

   $ python manage.py makemigrations
   

3. Perform Migrations.

   $ python manage.py migrate
   

4. Create a super user to access administrator controls and dashboard.

   $ python manage.py createsuperuser
   

5. Run the program.

   $ python manage.py runserver
   
   or
   
   $ python manage.py runserver --nothreading --noreload
   

6. Make sure to navigate in your browser to http://localhost:8000/ to view the application.

7. To access the admin type the following into the browser while the server is running http://localhost:8000/admin. This page will display the database and allow the admin to edit user's accounts.

8. To access the register type, please write an username and upload a image of this user to register into the system.

The images taken in registration process are saved in \login_recognition_app\facialrecognition\media\images --> These are our base images for the recognition process.

9. To login into the system, please click on Login, write the registered username and then look the camera for the login photo verification process.

Then, press Q for take the picture for login.

If the detected face make match with a registered username, then you'll have a Welcome like this:

10. Press CTRL-C to stop the process.