FACE MASK AND SOCIAL DISTANCE DETECTOR USING OPEN CV, DEEP NEURAL NETWORK, AND YOLO OBJECT DETECTION

Abstract

The COVID-19 pandemic is a global pandemic that becomes the highlight of the year 2020 and is still ongoing in early 2021. COVID-19 transmission is very rapid and easy, simple preventive action such as wearing a mask and do physical distancing should always be done to be protected from COVID-19. In public places such as stores, offices, airports, train stations, etc., it is very difficult to control the obedience of the people to always carry out the COVID-19 health protocols and it is a risk to place a guard/security in such places since they can be easily infected. This project will provide a desktop application that can monitor the COVID-19 health protocols in public places especially for face mask usage and social distance violation. The programming language used to develop this computer vision application is Python along with the OpenCV library. Face and face mask detection use the DNN library to train the dataset. Distance detection uses YOLO object detection to train the dataset model for people detection. The detection result used for monitoring is displayed on the windows page that was made using Tkinter. This project software is compiled using Visual Studio Code for the desktop application. Therefore, with the development of this project as in creating face masks and social distance detectors, it is hoped that health protocols can always be applied, especially in public places to stop COVID-19 transmission.