The third covid-19 wave has once again brought us back to the beginning. However, we are now learned and experienced. The other day I was at a local mall and the social distance monitoring system only consisted of a single watchman asking people not to crowd.

What it does

The software uses a live video feed to count the number of people in the video frame. Moreover, it counts the number of people of who are in at high risk. Once the number of people at high risk are above a threshold an automatic message will be sent to a phone number using the Twillio API.

How we built it

We used Python, Open CV, and Yolo object detection.

Challenges we ran into

We tried to make our application usable via any RTSP protocol wireless camera. We planned to stream live video using the RTSP video on a Flask website. However, we ran into errors while live streaming the video and using object detection in flask.

What we learned

I had never used Yolo object detection before so that was a learning curve.

What's next for Social Distancing Analyser

We plan to create a working website where any RTSP camera can be linked to make mass social distancing accessible.

Built With

Share this project: