Design and development of a computer-vision-based robotic arm system for sorting of vehicle-like objects
Abstract
The application of computer vision in vehicle-like object sorting systems contributes to improving productivity, accuracy, and reducing human-induced errors in industrial production lines. By exploiting visual features and combining image-processing techniques, deep learning enables reliable and automated recognition of complex objects under varying lighting conditions, viewing angles, and working distances. This paper presents the design and implementation of an automated sorting system integrating a robotic arm with a YOLO-based deep learning model for real-time detection and classification of three vehicle categories—cars, trucks, and buses—while extending classification capability through color recognition and QR code identification. The system supports three recognition modes: YOLO combined with color detection, QR code–bas ed recognition, and a hybrid approach. Recognition results are transmitted to a Siemens S7-1200 PLC to control the robotic arm, while monitoring and operation are performed via a SCADA interface. Experimental results demonstrate that the proposed system operates stably in real time and achieves high classification accuracy under different working conditions, confirming the feasibility and effectiveness of integrating deep learning–based computer vision with PLC control for industrial automation systems.