An inexpensive stereoscopic vision system for robots

Abstract

A relatively low cost stereoscopic vision system has been constructed here at the University of Alabama in Huntsville, using a video camera interfaced to a microcomputer. A motor driven mirror assembly attached to the camera is used to obtain the stereoscopic images. In the preliminary studies, an APPLE II microcomputer fitted with a DITHERTIZER II board was used for digitizing the video images. With its 280 × 192 resolution, we are able to measure distances with an accuracy of 8%. Currently, an IBM PC fitted with a VAN GOGH board by TECMAR having a resolution of 256 by 256 is used in place of the APPLE system. Using this system, much better results are obtained. The end system will be integrated to a Cincinnati Milacron T <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> industrial robot. In order to meaningfully utilize the information provided by the various subsystems interfaced to the robot, a three level hierarchical distributive system is being set up, using a PDP 11/34 minicomputer as the supervisory computer.