CodeVsCOVID19 - Minimal Distance
In these trying pandemic times, everyone needs to do their part to help aid the overall public health and safety. One way in which the WHO and CDC has recommended as a very effective method for doing this is social distancing. While social distancing is effective, people often don’t adhere to the guidelines that define proper social distancing. So, we created Minimal Distance.
What it does
Minimal distances uses IOT and Computer Vision to enforce social distancing guidelines. The IOT part tracks when bluetooth signals enter a “zone of importance,” aka a building. Once a capacity in which the building is no longer safe to be in is reached, the doors to the building lock for the outside until someone leaves. The computer vision aspect of the project uses existing CCTV cameras inside of these buildings for a live video stream and then analyzes these videos using advanced computer vision algorithms to determine if people are staying at least 6 feet apart.
How I built it
For this project we have many moving pieces that each member worked on. We developed a UI / UX on figma. We used a flask server with a computer vision library to track people movements in a room (though in a limited scale). Node and vue were used to build the frontend portion of the site.
Challenges I ran into
Being a remote hackathon, not being in-front of my teammates definitely made it harder to collaborate to the extent of which I had hoped. Nevertheless, through GroupMe, Discord, Zoom, and Slack we as a team overcame this challenge and worked hard to boost our overall collaboration effectiveness.
Accomplishments that I'm proud of
Creating a project that could be deployed into production today that can help society flourish in these trying times.
What I learned
How to better work in a remote environment using tools like Zoom and Slack.
What's next for Minimal Distance
We hope to continue to build on this project in the future and any advice is appreciated.
Try It out
computer-vision, css3, domain.com, flask, google-cloud, html5, mongodb, node.js, python, vue