Inspiration

Haloween involves dealing with pesky kids taking more sweets than they should. We want to treat the right kids with sweets without getting involved with the kuffufle.

What it does

It uses machine learning to identify a normal human face to a Halloween face/mask. If the person is indeed trick or treating then sweets will be dispensed through automated dispenser controlled by a raspberry pie module using a stepping motor.

How we built it

We split the team into two, hardware development interacting with the motor and structure of dispenser and the image processing team dealing with cameras and building facial recognition software to be uploaded to the pi

Challenges we ran into

Installing the correct packages and requirements. Configuring GPIO ports and the associated clocks

Accomplishments that we're proud of

It works Facial recognition Hardware operation Integration of hardware and software

What we learned

How to do facial recognition Embedded systems Real life application

What's next for Scary Face Recognition Sweet Dispenser

A moving robot that can identify kids Deep learning so the system can be fully autonomous A system that learns and adapt autonomously

Built With

Share this project:

Updates