Advanced Customer Behavior Tracking and Heatmap Analysis with YOLOv5 and DeepSORT in Retail Environment

Abstract

This paper presents a computer-vision-based approach designed to enhance product placement and sales strategies in physical retail stores through real-time analysis of customer behavior. Our method employs DeepSORT for tracking and YOLOv5 for object identification to generate heatmaps that illustrate consumer movement patterns and engagement levels across various retail locations. To precisely track customer paths, the procedure starts with the collection of video material, which is then analyzed. Customer interaction and traffic patterns across various retail zones are represented using heatmap visualization, which offers useful information about consumer preferences and product popularity. In order to maximize customer engagement and optimize the shopping experience, businesses may use the findings of this analysis to improve product placements, store layouts, and marketing strategies. With its low intervention requirements and scalable and non-intrusive solution, this system may be used in a variety of retail environments. This system offers a scalable and non-intrusive solution that requires minimal intervention, making it adaptable across different retail settings. Our findings demonstrate the approach’s effectiveness in identifying strategic areas for improvement and adapting retail environments based on real-time customer interaction data. This study underscores the potential of computer vision in retail analytics, enabling data-driven decisions that enhance both customer satisfaction and operational efficiency. This approach gives merchants useful data to develop more responsive, customized, and effective shopping experiences by providing a dynamic perspective of consumer behavior. Retailers may promote a modernized and customer-centered retail management strategy by using this creative application of computer vision to match marketing tactics and shop design with real consumer behaviors.