In-store Intelligent Customer Counting and Monitoring System

Abstract

A big group of individuals that have congregated in one location is referred to as a crowd. The relevance of crowd surveillance is becoming more apparent to different security and event management organizations all over the globe as a result of the rising worry about the general population. These difficulties with crowding seen in retail malls need immediate attention. This article will discuss how we are utilizing picture classification and object recognition to acquire a count of the number of individuals currently in the shop. A sophisticated crowd monitoring system for the shop has been designed by our team employing the regression techniques. In this case, we utilize the Yolo algorithm for real-time object identification. Yolo is a method that employs a single neural network for the purpose of object detection. To put it simply, we want to anticipate a set of objects and the bounding box that specifies where the objects are located. Each box may be traced by employing a total of four descriptors, including the center of a bounding box, its height, and its breadth value, which each relate to a different grouping of objects. It is anticipated that this would boost the ability of shops and consumers alike to gather information about traffic patterns of individuals. Because of this, we are able to make predictions in almost real-time, and it is simple to obtain information such as the density map, the calculation of the number of people or count, and the retail rush hour.