Intelligent Drone-Based Framework for Autonomous Inventory Data Inspection

• A drone-based system for autonomous inventory inspection using deep learning (DL). • UAVs autonomously captured inventory images; DL models detected barcodes in images. • Achieved detection rates up to 100% and decoding accuracy between 81.3% and 98.6%. • The system shows strong applicability for industrial and commercial inventory tasks. This study introduces an innovative drone-based framework for autonomous inventory inspection, designed to address the inherent inefficiencies and limitations of conventional inventory management practices. The proposed system uniquely combines indoor drone navigation with advanced deep learning-based barcode detection and decoding, further enhanced by the creation of a manually annotated barcode dataset acquired directly from drone imagery in warehouse environments. In contrast to previous studies, which have typically examined these technological components in isolation, this research presents a comprehensive and unified approach that facilitates efficient, contactless, and dependable inventory data collection. State-of-the-art deep learning models, specifically EfficientDet, Faster R-CNN, and YOLO, are utilized to accurately identify and decode both one-dimensional and two-dimensional barcodes, thereby streamlining inventory identification and data acquisition processes. Experimental evaluations conducted under controlled warehouse conditions have demonstrated robust performance, achieving barcode detection rates of up to 100 percent, with a minimum of 74.1 percent, and decoding accuracy ranging from 81.3 percent to 98.6 percent. The primary contributions of this work include: (1) the development of the first unified framework that integrates indoor drone navigation with DL-based barcode reading in real-world settings, and (2) the creation of a new annotated barcode dataset tailored for warehouse environments. These results highlight the significant potential of this approach to transform inventory management practices and enable scalable, efficient, and autonomous operations for industrial and commercial applications.