Inspiration

Protecting lives by detecting tired and distracted driving using cheap, effective tools.

What it does

EyeO tracks drivers' eyes to detect when they are falling asleep. If it detects a driver falling asleep, it sounds a buzzer and flashes a light to wake the driver. The device is small enough to hang from a car visor.

How we built it

EyeO was built by combining OpenCV, a computer vision library, with Dlib, a machine learning library. By taking video of a driver's face, we can determine the aspect ratio of open vs. closed eyes. If the driver's eyes are closed for too long, we sound an alarm.

EyeO runs on a raspberry pi and uses a web camera, so it would be cheap to manufacture and add to any car.

Challenges we ran into

Dependencies are hell. The raspberry pi has limited computational power. Lack of prior experience with C++.

Accomplishments that we're proud of

Jumping into a new, complicated technology stack and creating a working product in 36 hours.

What we learned

How to use OpenCV and Dlib. Patience.

What's next for EyeO

Recording drivers' overall wakefulness and uploading it to a web site so they can track their driving habits. Tracking driver wakefullness provides another metric for insurance companies that use driver behavior tracking devices.

Built With

Share this project:

Updates