Computer Vision/IoT project for HackUCI 2020
Skate safer with SafeSkate!
Real-time detection of obstructions for skareboards via computer vision
Alerts skater of objects on the road that may cause a fall
Are you a skater?
If yes, then chances are you have probably fallen from obstructions on the road.
What if something could warn you of objects getting in your path and save you from a potential fall?
That's where our project SafeSkate comes in.
Using a camera mounted to a skateboard, live footage is collected as you skate.
This footage is analyzed in real-time using computer vision via OpenCV on a Raspberry Pi.
If any object large enough to interfere your skating is detected, the backend script updates a boolean value on Firebase to denote that an obstruction exists.
Our frontend (we hope to integrate this into the remote for electric skateboards, but for now ran on your phone) continuously fetches from this database and warns the user with a beeping sound if an obstruction is detected.
Object detection is done with recognizing contours. Contours can be explained simply as a curve joining all the continuous points (along the boundary), having same color or intensity.
Our backend script processes the video footage in real-time looking for contours to detect objects in the way.
The images below show what the camera is seeing and detecting. Normal(left) vs Contour Detection(right)
No object detected
The frontend changes when something is detected (Beeping simultaneously)
Backend:
- Python
- OpenCV
- numpy
- Firebase
- Raspberry Pi
Frontend:
- React.js
- HTML/CSS
pip install -r requirements.txt