Webcam-Based Physical Distancing Prototype to Calculate Real-Time Distance by Applying Euclidean Distance

Abstract

The Covid-19 virus has spread rapidly in Indonesia in recent years, significantly increasing the number of victims in the community. However, there is a lack of warnings for violators of physical distancing. This study aims to create a system as a solution to overcome the problems of implementing physical distancing protocols. Image capture media in the form of cameras were used to monitor human objects in crowded situations by tracking the movement of objects at regular physical distances. In this study, computer vision, image processing, and pattern recognition were used to construct the distance measurement between humans. Using a webcam by applying the Euclidean Distance and Haar Cascade methods, this system is intended to determine the distance between human-made objects. The results demonstrated that a digital image system employing the Haar Cascade and Eucludien Distance methods could make it easier to control the gathering point in a given area. Researcher has developed a software usage assistance feature for use with this application. Based on the findings of the study, the accuracy of this system was evaluated using 100 trials, with a succes rate of 100% since the system resulted correct notification in all trials. In addition, this digital image system is anticipated to assist developers and the general public in controlling gathering locations that do not adhere to established health protocols, allowing violations of physical separation to recognize their errors when captured on camera immediately.