Share Email Print
cover

Proceedings Paper

Guidance Of A Mobile Robot Using An Omnidirectional Vision Navigation System
Author(s): Sung Jun Oh; Ernest L. Hall
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Navigation and visual guidance are key topics in the design of a mobile robot. Omnidirectional vision using a very wide angle or fisheye lens provides a hemispherical view at a single instant that permits target location without mechanical scanning. The inherent image distortion with this view and the numerical errors accumulated from vision components can be corrected to provide accurate position determination for navigation and path control. The purpose of this paper is to present the experimental results and analyses of the imaging characteristics of the omnivision system including the design of robot-oriented experiments and the calibration of raw results. Errors less than one picture element on each axis were observed by testing the accuracy and repeatability of the experimental setup and the alignment between the robot and the sensor. Similar results were obtained for four different locations using corrected results of the linearity test between zenith angle and image location. Angular error of less than one degree and radial error of less than one Y picture element were observed at moderate relative speed. The significance of this work is that the experimental information and the test of coordinated operation of the equipment provide a greater understanding of the dynamic omnivision system characteristics, as well as insight into the evaluation and improvement of the prototype sensor for a mobile robot. Also, the calibration of the sensor is important, since the results provide a cornerstone for future developments. This sensor system is currently being developed for a robot lawn mower.

Paper Details

Date Published: 1 January 1987
PDF: 13 pages
Proc. SPIE 0852, Mobile Robots II, (1 January 1987); doi: 10.1117/12.968260
Show Author Affiliations
Sung Jun Oh, University of Cincinnati (United States)
Ernest L. Hall, University of Cincinnati (United States)


Published in SPIE Proceedings Vol. 0852:
Mobile Robots II
Wendell H. Chun; William J. Wolfe, Editor(s)

© SPIE. Terms of Use
Back to Top