Social Distancing Detector Using Deep Learning

A visual surveillance system is a tool used for analyzing and explaining object behaviors. A visual surveillance system typically includes the detection of static and moving objects. Video tracking is also used to better understand what’s happening in a scene, making it an important part of computer vision theory. This paper describes about how detecting and tracking humans is done and check whether they are following social distance or not. This paper uses the Yolo V3 object detection model, which segregates people and then attempts to determine whether they are adhering to social distance standards based on the bounding boxes that track them. This paper also shows the number of violations, and also how differently colored boxes bound around the people depending on the distance. To monitor and manage social distancing protocols, various technologies such as computer vision, deep learning approaches, and the programming language Python are often used. The motivation of this paper is to develop a method for identifying and informing officials in charge of breaches in social distance regulations in public areas.