Vision-guided automatic tasks applied to a 5-DOF robotic arm

Abstract

There is still a wide field to explore and develop in the robotic intervention branch, particularly in which concerns robots autonomy when interacting with the environment and human beings. Motivated by the benefits of the use of service robots and its ascension, this study aims to show the possibilities of using a robotic arm for automatic tasks. Presented experiments were driven using the PhantomX Reactor Robot Arm (an open-source robotic arm) within the ROS framework alongside computational vision applications using OpenCV and ArUco markers libraries. This setup made it possible for the robot to search and recognize markers using a camera attached to it and then manipulate objects under predefined movements. The results demonstrate the robot can well-perform tasks, being reliable and accurate, making it possible to scale up those functions on intervention industrial robots, as the concepts are the same and the functionalities are similar.