Inspiration
-Monitor lecture halls, and venues to make sure they are fire safe. If the room is too crowded or exits are blocked, it will alert security/fire department via text message.
What it does
Our project video streams from a camera or image input and analyzes the image(s) using YOLOv3 machine learning modules to detect people. Our program supports four different inputs: Twilio text messaging, Twilio image url, slack, and a camera. Using any of the four input options, our program will return the number of people detected in the room as well as whether or not the maximum capacity for the room has been exceeded.
How we built it
Written in Python using OpenCV with YOLOv3 object detection. Our user interface is Twillio text message api.
Challenges we ran into
It was challenging implementing many different supports for different inputs and making sure that each one was compatible with one another.
Accomplishments that we're proud of
We were able to get person recognition to work withOpenCV using laptop cameras (with a supported version for phone cameras). We implemented the Twilio API which allowed us to get updates in room occupancy through text messages to and from a server.
What we learned
Learned about computer vision with OpenCV. Dabbled in machine learning/object detection with YOLOv3.
What's next for
-Create an program where users can get live updates to how long lines are at their favorite restaurants. Utilizing twillio, people will be able to send texts to see how many people are in line.
Log in or sign up for Devpost to join the conversation.