Physical Distancing System Using Computer Vision

The regulation of social distancing in standing queues such as queues at ATMs and queues for ordering fast food is very important to apply during COVID-19 pandemic. The system can control the discipline of keeping a distance which can reduce the rate of transmission of the Covid 19 virus. The system is expected to be able to handle supervision and regulation of the application of physical distancing by involving hardware input in the form of a camera, which will provide the condition of standing queue in the form of images to be processed to produce output information, the number of objects in the standing queue, the distance between objects and the x, y coordinates of the object in the standing queue. The final result of the Physical Distancing Warning System is able to maximize the distance control in standing queues such as queues at ATMs and queues for ordering fast food. Based on tests carried out able to detect human objects and calculate the distance between them. Tests to obtain the validity of system performance are carried out by varying the minimum threshold value. The best performance results are obtained at the minimum threshold value of 40 where the system error is 22.75%. The output of the computer vision process that is sent to the microcontroller as a command to the lights, servo and laser has been fully worked (100%) with the correctness of the response completely controlled by the processing results in the computer vision section.