Social Distancing Violation Monitoring Using YOLO for Human Detection

Social distancing may be implemented however people are not self-aware in tracking their distances between other people. This study developed a device using Raspberry Pi that measures the distance between people for monitoring social distancing. The system used YOLO algorithm to detect humans within the given camera frame. It also used the Euclidean distance formula to calculate the distances between people detected. Then, violators were identified as people that are not observing at least 1 meter apart from other people. They are distinguished in the display by having red highlights while non-violators had a green highlight. The display was also divided into five areas to properly notify the violators, where LED was used to notify violators near the tables and an announcement was provided for violators at other areas. To control the LED wirelessly, ESP8266 was used to serve as a microcontroller and Wi-Fi module. Then, a paired T-test for the comparison of the computed and actual distance measurements yielded a T-score of 0.0714 which got a 0.94 probability value. The calculated probability value denotes no significant difference between the computed distance obtained from the device and the actual distance measurement from the controlled setup.