Smart Distancing

Abstract

In this paper, an integrated smart distancing that fuses mask detection, degree of store crowdedness and zero contact billing system is proposed. The proposed system acquires real time data and determines whether the person is wearing a mask or not utilizing the Machine Learning packages like TensorFlow, Keras, OpenCV and MobileNetV2 architecture. MobileNetV2 is fine tuned to implement COVID-19 face mask detector training script. Furthermore, With the help of A.I combined with edge computing, Object tracking and object detection the degree of store crowdedness can be measured and recorded in real time using OpenCV. An alert in the form of mail is triggered in case of breach of threshold. This obtained dynamic data is rendered in our customized website created using HTML, CSS and JavaScript. The intervals of this operation are monitored by a scheduler. Finally, a smart shopping system is designed by employing the concepts of RFID modules and tags integrated with NodeMCU alongside IoT. This system aims at developing an automated self- billing system that would save the customers time and helps the people abide by the rules of the pandemic.