Visual-based Guidance System for a 6-DOF Robot

Abstract

The idea of this bachelor’s thesis is to develop a position-based visual servoing system with two webcams that allows an anthropomorphic robot ABB IRB 120 with six degrees of freedom (DOF) to be guided in real time by an operator, by triangulating a specific target moved within the field of view (FOV) of a machine vision system. For this kind of motion control based on visual data input, termed visual servoing control, we need a stereo vision system to acquire the images of the target since three-dimensional information is required to perform object tracking with six DOF. These images are processed by a MATLAB ap-plication running in a remote PC. The current coordinates of the target referred to the left camera reference frame are extracted from the images and sent through an Ethernet connec-tion to the robot controller, which is programmed to receive the vectors and move its tool centre point (TCP) to the demanded position within its workspace.