GuardDog
Description as a Tweet:
GuardDog is a home security system. Each time an unrecognized person is detected by the camera, an email notification is sent to a selected address. It'll be able to turn on lights, or perform other actions as intended by the user.
Inspiration:
Guard dogs!
What it does:
GuardDog uses the YOLO system to detect an intruder in real time. Once caught by the camera, it will send an alert to the user, who can take next steps manually, or just set the system to turn on lights, etc.
How we built it:
Used the YOLOv3 (You Only Look Once), which is a real time object detection system. Wrote the backend using Python and OpenCV.
Technologies we used:
- Python
- Raspberry Pi
- AI/Machine Learning
Challenges we ran into:
Mainly trying to implement this project under the 36-hour time constraint.
Accomplishments we're proud of:
It is a very light-weight recognition system which will work on most microcontrollers.
What we've learned:
Learnt a lot about the YOLO system and OpenCV since I haven't used those before.
What's next:
Integrating all the features that I wasn't able to.
Built with:
Python, OpenCV, Raspberry Pi, Tensorflow, VSCode.
Prizes we're going for:
- Best Hardware Hack
- Best Documentation
- Best Security Hack
- Best AI/ML Hack