STRAIGHT-LINE MOTION CONTROL FOR AUTONOMOUS LAND VEHICLES USING ID IMAGE-PROCESSING TECHNIQUES

In this article, we develop a novel approach to detecting the movement changes of an autonomous land vehicle. Movement changes are of two pa n values, namely, the pan angle and the pan distance between the vehic le and the navigation line. Both of these two values have to be taken into account while making movement calibration. The approach is compri sed of a learning phase and an operation phase. in the former phase, a navigation map is generated by taking four images based on the preset maximum and minimum tolerant pan angles of the vehicle, while it is l ocated on the preset maximum and minimum tolerant pan distances on the side parallel to the navigation Line. In the latter phase, images are taken in regular time intervals while the vehicle follows the navigat ion Line, and the two pan values are calculated through interpolation based on the navigation map. To increase the performance of the propos ed system, two refinement approaches are also presented. Experimental results are given to reveal the superiority of the proposed approach. (C) 1998 John Wiley & Sons, Inc.