Inspiration
Primary parking at Missouri-St. Louis is a parking garage near the library. We were curious as to how many cars drove through the lot at certain times of day, when the peak traffic hours were, and how many individuals used the lot. To this end we wanted to develop an AI system that could answer these questions for us, and be flexible enough to be extensible in the future.
What it does
GarageView used an open source deep neural net system through caffemodel to recognize objects in a frame. We utilized this model with OpenCV and python to set up a Rasperry-Pi3 equipped with Picamera as a traffic comprehension unit.
How we built it
We used Python 2.7.10 with OpenCV and NumPy modules to run the object recognition models and a Rasperry-Pi3 with Picamera V2 as the hardware.
Challenges we ran into
We were unfamiliar with Raspberry-Pi and OpenCV when we began this project. Understanding them enough to utilize them was at first difficult...
Accomplishments that we're proud of
...but we managed to do so and learned new hardware and software systems that we didn't know before this hack.
What we learned
That to create something worthwhile, one must first deal with setbacks in order to gain knowledge.
What's next for GarageView - Vehicle Storage Tracking System
Network connectivity and a phone app.
Built With
- caffemodel
- computervision.dnn
- opencv
- picamera
- python
- raspberry-pi
Log in or sign up for Devpost to join the conversation.