<![CDATA[Manual sorting in manufacturing is time-consuming, labor-intensive, and prone to errors, especially when items have similar colors and shapes. This study aims to design and implement an automatic sorting system for goods based on color and shape to enhance production efficiency. The system integrates a webcam for image acquisition, the YOLO v5 object detection algorithm for real-time classification, and the Outseal PLC to control actuators via a Ladder Diagram. An experimental method was used, with a dataset of 18 object types tested under three lighting conditions (daylight, low light, and ring light). Performance was evaluated using a confusion matrix, achieving an average accuracy of 88.26% and precision of 70.38%, with the best results under ring light illumination. These findings demonstrate that the proposed system can reduce operational costs and improve productivity for small- to medium-scale industries. Future work should include extended field testing and adaptive algorithms for varying lighting environments.]]>
Copyrights © 2025
