Mask Detector

In light of recent events and the significant impact that COVID-19 has had on our daily lives, our team of 4 has decided to create an easily implementable technology that could better protect the public and reduce the spread of COVID-19.

We have created a facial recognition software to detect whether or not someone is properly wearing a face mask using machine learning algorithms. The associated actions could range from alerting a system operator, to opening automated doors, or prompting unmasked persons to put a mask on. There are numerous potential applications that may stem from this project, and we believe that the flexibility of our implementation could make a significant impact on public health.

Hardware

Raspberry Pi 3B
Raspberry Pi Camera
Power Supply

Software

Linux/Ubuntu
Raspbian OS
Python3, PIP
Git/Github Machine Teacher:
teachablemachine.withgoogle.com/

Packages/Libraries

Tensorflow.keras
OpenCV
numPy
Pillow
PIL
RPi.GPIO
PiCamera

How It Works

When activated, our Raspberry PI 3B will continuously take pictures of the person in question, going through several steps to determine if a mask is being worn properly.

Using OpenCV's integrated facial recognition algorithms, we first determine if there is a human face(s) present in the view
Upon recognizing a human face, the program then checks if you have a mask on by using a TensorFlow machine learning model that we trained.
If the program determines that the person is not wearing a mask, you will be alerted by several blinking red LEDs. If a mask is being worn correctly, then green LEDs will blink instead.

Real Life Application

Can be used in a variety of situations to improve personal safety, hygiene, and adherence to regulations. Potential applications range from personal use to commercial implementations. Some usages could include:

Store fronts / Public Areas
- Could be connected to an automated gate, removing need for human moderation
House / Apartment Buildings
- Protect yourself and delivery personnel
- Could set up reminder in case you forget to wear a mask
- Could be further developed into facial recognition to unlock door
Schools and Hospitals
- Ensure students, patients, visitors are properly protected
- Reduce spread in high-risk areas such as hospitals, schools with young children