Intelligent Retail Store Monitoring Using YOLO+AI

Abstract

Store businesses face a persistent challenge in ensuring that shelves are adequately stocked and products are consistently available for customers on racks. Manual checks are inefficient and often lead to delays in stock refilling, resulting in customer dissatisfaction and potential loss of sales. To address this issue, we propose an AI-based smart monitoring system designed for real-time detection of products displayed on shelves. This solution is a cost-effective design that can be deployed as an on-premise system. It utilizes a YOLO model trained on customized datasets to detect products and classify them into three categories: in stock, low stock, and out of stock. This classification triggers timely alert notifications to staff, enabling faster restocking and improved shelf management efficiency. The system is scalable and can be easily integrated with inventory management dashboards. It ensures reduced operational costs while significantly improving inventory tracking. By automating shelf monitoring, the solution replaces traditional manual methods with intelligent AI-driven stock-check systems, ensuring timely product replenishment and minimizing delays. The system includes real-time image acquisition and YOLO-based inference along with a robust data collection and preprocessing module to ensure high-quality input for model training and deployment. Captured images undergo preprocessing steps such as resizing, normalization, and augmentation to enhance model accuracy before being stored in a centralized database. A decision engine categorizes items into in-stock, low-stock, and out-of-stock based on predefined thresholds. The system generates multi-modal alerts through visual displays on kiosks, voice notifications for staff, and email alerts for managers. This closed-loop system supports proactive shelf replenishment, reduces stockout incidents, and improves overall customer satisfaction and operational efficiency.