Infectious disease is a terrible pandemic around the world, the economy and even people's lifestyles are affected while implementing a lot of standard operating procedures (SOPs) to control the spread of the virus. One of the most effective ways is to control the physical distance between people. Therefore, a Social Distancing Monitoring (SDM) system can be used to track the safe distance between people. In this study, a comparison between algorithms will be made to identify the most efficient algorithm based on human detection results. As a result, the OpenPose algorithm has outstanding human detection results compared with YOLOv4 and YOLOv4-pedestrian. Furthermore, a distinctive bounding box method for human detection has been proposed, which is defined from the human skeleton using the OpenPose algorithm. Hence, SDM will be efficient after a perspective transformation to calculate the distance between people, since most of the people in the frames have been detected. Finally, this study will discuss how the power of AI enhances the SDM system in detecting humans and calculating the distance between people detected.
