Abstract
This paper presents about a solution for social distancing using computer vision technology, especially in a queue scenario. Social distancing is a series of measures used to prevent the spread of infectious diseases by maintaining physical distance between one person and another. To overcome these problems, we need a monitoring system that is used to detect violations in line. The system will detect the distance between people in line by using image processing. If the distance is not as recommended shorter than the limit, then the system will issue a warning through the output of image processing on the monitor. With this warning system, visitors are expected to comply with existing regulations. In this experiment, we compare two object detection algorithms. That algorithms are YOLOv3 (You Only Live Once) and mobilenetSSD (Single Shot Detector). Additionally, both algorithms are also implemented in CPU (Central Processing Unit) and GPU (Graphics Processing Unit) platforms. Based on our experiments, YOLOv3 on the GPU platform offer the best performance in term of accuracy and speed.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
