Social Distance Monitoring using Deep learning

Abstract

Disease transmission can occur in a variety of ways, but the most unexpected way is through air transmission. There are very few diseases that can be spread through the air. Dust particles and respiratory droplets carry airborne pathogens, which are finally dissolved by others. An airborne disease can be diagnosed without being in the same room as a sick individual. In December 2019 the covid-19 was identified. Social distancing one from the disease is an effective way to slow down. In the existing crisis. The population's vulnerability is exacerbated by the lack of effective treatment medicines and immunity for COVID-19. There are some vaccines available in the market but there is no permanent way, social distancing is the only way to fight against those diseases. Nowadays research in deep learning fields is going on in every industry for better accuracy and less involvement. So, this chapter is mainly focused on monitoring the people who maintain the social distancing from the video using pre-train YOLOv4 algorithm which is fast. YOLOv4 algorithm is useful to detect multi objects from the real time videos. So, using deep learning algorithms the chance of spreading the virus and air disease will be reduced. In this chapter, we are going to utilize the YOLOv4 algorithm and detect multiple humans and calculate the distance between two objects and if two humans are violating the rules of social distancing as per the government rules then the alert will be triggered. We also compare the predefined YOLOv4 model confidence score with the YOLOv3 model for better accuracy.