Digital Solution for Enforcing Social Distancing
5 Pages Posted: 2 Jun 2020 Last revised: 20 Oct 2021
Date Written: May 31, 2020
During the period of pandemic, it is very crucial to follow the instructions given by oﬃcial authorities. One major step during this period is the Social Distancing. It becomes very diﬃcult for the authorities to keep detailed check on whether social distancing is being followed or not. We come across situations where instead of patiently following these rules, people panic and tend to crowd up places. So, for easy monitoring for the authorities (like police oﬃcials) we propose to you a digital solution using Deep Learning technique which would alert them as soon as the violation of the social distancing is detected that is number of people more than the threshold (limit on the number of people allowed to be in a place, set by the government) or distance between two people is less than the threshold distance. A video stream will be captured from the CCTV camera, with the help of PoseNet model we are detecting the humans and keeping a track of the number of humans present in the given live video stream, if the number of humans crosses the minimum threshold limit (set by the oﬃcials) or if the Euclidean distance between any two poses detected in the frame in less than say 3ft we alert the authorities in-charged. This application will not only be helpful to reduce the eﬀorts and strength required to manually check the places around, but also to save time and quick analysis as in layman’s term the CCTV cameras will help simultaneously monitor each and every place of common gathering.
Keywords: pandemic, deep learning, PoseNet, human detection, alert
Suggested Citation: Suggested Citation