In the car system
This was inspired by the increasing number of accidents caused due to fatigued drivers, especially truck drivers. Any long distance drivers is at the risk of falling asleep behind the wheel.
What it does
Our module detects if eyes have been closed for a "long" period of time and then evoke the speaker module to emit a wake up call ("WAKE UP SLEEPYHEAD"). The camera moves along with your face alignment.
How I built it
The project was divided into the software and hardware components
One person worked on the software and was involved with monitoring a stream for faces, detecting drowsy eyes using the concept of eye aspect ratio, and then setup the alarm sound.
The hardware part was challenging. We used a Raspberry Pi 3 with a servo motor to mount the USB webcam. After the wifi was enabled, we use the faces detected from the CV algorithm to keep the person of interest POI at the center of the frame.
Challenges I ran into
The wifi module of our raspberry pi refused to work for the initial 5 hours of the hackathon. We finally managed to mount a new image and got it running. The project was challenging due to running the heavy deep learning library on the RPi.
Accomplishments that I'm proud of
We managed to have a running system of an RPi that can be mounted on your car and it can track your face movement also
What I learned
RPi is a great computer hack when the wifi module works!
What's next for SleepyHead
We want to incorporate a voice assistant capability to it. For now, we ran out of time.