Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Acquisition of 3-D Information by Binocular Stereo Vision for Vehicle Navigation through an Orchard

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  Pp. 337-346 in Automation Technology for Off-Road Equipment, Proceedings of the July 26-27, 2002 Conference (Chicago, Illinois, USA)  701P0502.(doi:10.13031/2013.10023)
Authors:   Teruo TAKAHASHI, Shuhuai ZHANG, Hiroshi FUKUCHI
Keywords:   Stereo vision, 3-D location, Image processing, Vehicle navigation

The purpose of this study is to establish a method of 3-D measurement by binocular stereo vision to obtain navigation information for an autonomous vehicle in an orchard. A method of image processing to compose a pair of images by left and right cameras was applied to reduce correspondence problems with stereo vision. The method is based on the concept of gaze and clearness of composite images analogous to the human visual system. Experiments on a trial stereo system were conducted using such objects as apple fruits and trees, cars, and persons in an orchard. The results showed that, when the width of an object image was not too wide (regardless of the object’s shape), and when color feature and variance of color density in the object differed from those of the surroundings, errors in distance measurement ranged from –10% to 5% within a distance of 20 m. The error enlarged for images with wide pixel areas, because false images appeared in unsuitable correspondence with the stereo pair. In conclusion, this method of binocular stereo vision is useful for obtaining 3-D information for a vehicle, though it is necessary to improve both the calculation of average variance of color density and the speed of image processing.

(Download PDF)    (Export to EndNotes)