Bryce Eldridge's MS Thesis Abstract

Vision-Based Tool Calibration for Industrial Robots

MS, Colorado State University, May 2006

Major Professor: Anthony A. Maciejewski

As offline programming of industrial robotic systems becomes more prevalent, the need for accurate calibration techniques for components in an industrial robot cell increases. One of the factors required for successful offline programming is an accurate calibration of the robotís tool frame. Excessive errors in the calibration of the tool frame will result in tool positioning errors that may render the system useless. Current methods for calibrating the tool frame are manual, time consuming, and often require a skilled operator. This thesis addresses the development of a simple, fast, vision-based method for calibrating the tool frame of an industrial robot.

Calibrating the tool frame is first examined from a mathematical perspective, in order to gain some insight into the nature of the problem. Then the basics of computer vision systems are reviewed, and a short discussion of camera calibration techniques is presented. Simulations of the vision-based tool calibration algorithm are then presented in both two and three dimensions, followed by a summary of the results obtained by testing the algorithm with an actual robotic welding system.

The tool calibration algorithm presented here offers several advantages over the methods currently in use. First, the vision-based method is significantly faster than current methods while delivering comparable accuracy. Second, the presented method requires minimal calibration and setup. Third, the method is non-invasive, i.e., requires no contact with the tool, and does not use any special hardware other than a camera, enclosure, and associated image acquisition hardware. While vision-based methods are not appropriate for every situation, using them to calibrate the tool of an industrial robot offers a fast and accurate way of linking the offline programming environment to the real world.