SafeMaps

SafeMaps!
Distance Detection
A view of the heatmap generated

Inspiration

Covid19 pandemic
Traffic maps used by news agencies.

What it does

SafeMaps is a platform that consists of a network of cameras installed in public spaces like parks and busy streets. Using computer visions, these cameras allow us to count the number of social distancing violations (maintaining a 6 feet distance between people) in real-time and log them in a database. This data is then displayed by a website that a user could check before making plans to determine the best places to enjoy the outdoors while also reducing the risk of exposure to COVID.

How we built it

We used OpenCV to process video inputs and send over the number of social distancing violations to a postgres db. This data is queried by the website and displayed to the user in a heatmap.

Challenges we ran into

Accounting for camera perspective in distance calculations
Working with the google-maps-react API to get the heatmap displayed

Accomplishments that we're proud of

Successfully implementing human detection and perspective correction algorithm
Putting together a simple but effective UI that displays all of the data in real-time as its collected

What we learned

A lot about computer vision techniques
Using google cloud API to implement the heatmap

What's next for SafeMaps

Running the OpenCV collection scripts on Raspberry Pis or actual cameras and running through a real world scenerio
Implementing autocomplete/search features on the heatmap