Using Computer Vision To Aid In Search and Rescue

Sample with One Person In The Frame
Sample with Two People In The Frame
Sample with Four People In The Frame
Tello Drone

Inspiration

There have been many accidents where counting people is useful in emergency situations. We wanted to build something that could aid in that. For example this technology could be used for rescue at sea if something like a large cruise ship hits something. A drone could be sent to accurately estimate the amount of people at sea and call in the appropriate rescue measures. It could also be used for accountability and making sure everyone is rescued and no one is left behind.

What it does

It uses OpenCV to take in the feed from the camera to figure out how many people are there. We used a third party model (YOLO Model). Live feed from something like a Tello Drone can be used to feed in information

How I built it

We build it using OpenCV and the YOLO Model. We made the program in Python.

Challenges I ran into

We were not able to get the live feed from the drone because of the issues with our computer systems and not being able to install certain Python dependencies. We used a webcam image instead to show that our program works. Also our PC was not powerful enough to run the YOLO model fast enough.

Accomplishments that I'm proud of

We are proud that we were able to finish the model and having it output the amount of people there are in the image. We were also able to launch the drone in the air and control it using our laptops.