RoboSort: Automated Object Sorting Robotic Arm

Abstract

This paper discloses the development and functionalities of an automated robotic arm system that eases the need for heightened precision and efficiency when sorting different objects. This is accomplished utilizing a 5-degree-of-freedom robotic arm, which reduces unnecessary handwork while increasing sorting accuracy and speed. It utilizes advanced computer vision and image processing methods based on the OpenCV library and Python programming to implement analysis of an object’s live video feed and object image capture. These images were implemented in the real-time webcam and processed through YOLOv5 for effective object detection, subsequent classification and sorting based on objects’ color and shapes using robotic arm. This method facilitates faster operations with closer to zero errors than humans, especially when critical precision is a requirement.