Jetson Nano and Arduino-Based Robot for Physical Distancing using Yolov4 Algorithm with Thermal Scanner

摘要

Coronavirus disease, more famously known as COVID-19, was first discovered in Wuhan, China; it was declared a global pandemic by WHO in March 2020. Due to the threatening characteristics of the virus, certain precautions had to be imposed by the government and health authorities to put the situation under control. To mitigate the further transmission of the virus, the "New Normal" was introduced to the public. This is by practicing the minimum safety protocols: wearing a facemask, frequently washing hands, and observing physical distancing. This study aims to build an autonomous robot that can monitor physical distancing, specifically focusing on people's queues. The robot utilizes the YOLOV4 Algorithm to detect the individuals and determine their Euclidean distance to determine if these people are observing the distance safety protocol that is 1.5 meters apart. The robot also includes a voice alarm that apprehends violators and reminds them to follow the practice. Moreover, the robot has an additional feature of detecting the body temperature of the people detected by the program. In assessing the robot's program, the implemented object detection achieved an accuracy of 93%, a precision of 87.5%, an error rate of 7%, and a recall of 94.6%. Moreover, by determining the constraint distance of the robot, which is 3.5 meters, the physical distancing program obtained a percent error of 4.26%.