Abstract: The COVID-19 (Corona virus) has sparked global panic after spreading to over 180 countries and causing 3,519,901 confirmed cases compared to 247,630 passing globally as of May 2020.The lack of any complex helpful experts, and therefore the need for resistance to COVID19, increases the population's defenselessness. Since there are no medicine available to cure, social distancing is the only viable option for combating this infection.Pre-trained deep neural network models, such as YOLOv3, is being used in this research to detect people with the help of bounding boxes that identify the human beings. The system also analyses the distances amongst people in the society using the Euclidean distance metric approach to estimate the proportion of people who infringe the social distancing within the camera footage.

Keywords: COVID-19, YOLO Algorithm, Video surveillance, Social distancing, Pedestrian detection, Pedestrian tracking.

| DOI: 10.17148/ IARJSET.2021.8587