Robotic Object Identification and Sorting Supported by Artificial Intelligence
摘要
The aim of this paper is to create an advanced system for identifying and sorting objects using the DobotTM MG400 robotic platform and artificial intelligence. This system aims to automate the process of classifying objects based on their shape and color, using the data obtained from DobotTM Vision Studio – an image processing software with a camera. The robot is capable of accurately classifying and placing objects in their respective positions, while the operator’s role is limited to initial commands such as start, stop, and reset. The results achieved demonstrate that the robotic system is effective for an automated industrial line. Shape identification has been successfully realized with an accuracy of 95% and color identification with an accuracy of 75%, respectively, based on the lighting conditions or changes in light for a brief time. Additionally, the system remembers occupied positions to ensure the correct placement of other parts in the available positions. Ultimately, this system can assist in industries where automatic classification and sorting of objects are necessary, reducing the need for human intervention and improving the efficiency of production lines