Counting Customers: An In-store Intelligent Customer Tracking and Monitoring System for Enhanced Retail Analytics

Abstract

A crowd is a large collection of people who have gathered together in one place. Because of the growing public concern, the value of crowd surveillance is becoming more clear to various security and event management organisations throughout the world. There needs to be rapid attention given to the issues with crowding in shopping malls. This post will go over how we are classifying images and identifying objects to get a count of how many people are in the store right now. Our team has used regression techniques to create a comprehensive crowd monitoring system for the business. Here, we use the Yolo algorithm to identify objects in real-time. Yolo is a technique for object detection that uses a single neural network. Simply expressed, we are interested in anticipating a set of items and the bounding box that defines their locations. The centre of a bounding box, its height, and its breadth value, which each pertain to a separate grouping of items, can be used as four descriptors in total to trace each box. It is believed that this would improve the capacity of businesses and customers to learn more about people's traffic patterns. As a result, it is simple to get information such as the density map, the calculation of the population or count, and the retail rush hour. As a result, we are able to make forecasts in virtually real-time.