The Effect of Distance, Viewing Angle and Static Time in Drone Image Detection Technology for Industrial Warehouse Inventory Scanning System

Abstract

The rapid growth of automation and Industry 4.0 has increased the demand for efficient inventory management systems in industrial environments. Manual inventory inspection is time-consuming, labor-intensive, and prone to human error, particularly in large warehouses and high-storage facilities. The objective of this study is to investigate the effect of distance, viewing angle, and static time in drone image detection technology for industrial warehouse inventory scanning systems. A low-cost quadcopter platform (DJI Tello) was employed, equipped with a camera system and controlled using Python-based programming. Image processing and barcode detection were performed using OpenCV and YOLO integrated with Pyzbar, enabling real-time identification and decoding of inventory labels. Experimental evaluations were conducted to analyze the effects of distance, viewing angle, and time-based static positioning on detection accuracy. Repeated experimental procedures showed results that demonstrated that the drone was able to reliably detect and decode barcodes within an optimal range of 30–50 cm, with 30 cm being the most consistent. The optimally detectable viewing angle is within 10 degrees of elevation, with 0 being the most consistent. The optimal time to be exposed to the barcode exceeds 1 second, with 2 seconds showing consistently adequate detectability. The findings confirm that UAV-based image detection systems offer a practical, cost-effective, and scalable solution for automated inventory monitoring. This study emphasizes the possibility of integrating drone technology with computer vision to enhance efficiency, accuracy, and safety in industrial inventory management applications.