Automatıc Social Dıstancıng Mechanism for Preventing Covid-19 Using Deep Learnıng

Abstract

COVID-19 is one of the viruses that was drastically spread, causing chaos in the entire world. Most of the scientists have studied the mechanisms to prevent this virus, and one of the mechanisms to mitigate the effect of COVID-19 is social distancing. People intentionally or unintentionally forget this distancing mechanism. We propose a novel mechanism, which will alert the users when they are approaching the others with a distance of fewer than two meters. This mechanism is available made available in an app where people can easily install this in the smart mobile. We have also developed a mechanism that automatically processes the CC tv footage and monitor the social distancing practice & alert the administrators. The system was thoroughly trained and tested with 20,410 videos and have achieved an average accuracy of 79.3%.