Machine Vision Surveillance System - Artificial Intelligence For Covid-19 Norms
Ritika Nigam, Arpan Kundu, Xinrui Yu, Jafar Saniie · 2022
This paper presents the design and implementation of the Machine Vision Surveillance System Artificial Intelligence (MaViSS-AI) for Covid-19 Norms using jetson nano. This system is designed to be cost-effective, accurate, efficient, and secure. The proposed system tracks and counts humans for monitoring social distancing and detects face masks using object detection methods. We used YOLO as an object detection method and neural network to detect a person and count them. And for social distancing monitoring the concept of the centroid is based on calculating the distance between pairs of centroids, and thus checking whether there are any violations of the threshold or not. To detect the face mask, a YOLO V4 deep learning model is used as the mask detection algorithm. The system also raises alerts when any suspicious event occurs. Given this alert, security personnel can take relevant actions. This research aims to provide a holistic approach to overcoming the real-time challenges encountered during the monitoring of Covid-19 norms.