Survey on Detection of Face Mask and Social Distancing using Tensorflow

Abstract

The global COVID-19 epidemic has had a significant impact on the entire world, infecting more than eight million people worldwide. The two most important safety measures that can be taken by the people in puplic in order to prevent this disease is wearing face masks and following social distancing. In order to create and keep an eye on this type of environment, we put forward an efficient and convenient CV(ComputerVision)-based approach mainly focusing on the automated real-time surveillance system to identify face masks, voilation and maintain social distancing in all public areas by creating a model on a system that monitors all activity and manages violations using a camera that can be a webcam, a phone, or even a networked device. After Detecting a violation, the system is able to send an alert popup/alert box including real-time pictures of the group or an individual breaching the basic COVID-19 Protocols to a privacy group of Telegram that was made for these type of cause. In this paper we have proposed a system which is a mixture of both modern deep learning algorithm and geometric techniques for making a model satisfying robustness condition. The proposed model is totally based on three aspects detection,validation and report. Hence.this model can solve a problem by allowing us to use it and save time and make the most out of it to decrease the spread of the pandemic and by reducing the total COVID-19 cases. It can be successfully implemented in present scenarios where strong public safety measures are required, such as public meetings, shopping malls, movies, grocery stores, and so on. Since the goal is to minimize the time taken and also the cost by decreasing the number of individuals to inspect in public places at a time,we make use of Automated inspection.