Road recognition for a wheeled heavy duty off-road autonomous vehicle

In this paper, an automatic control system based on "x-by-wire" is introduced for a wheeled heavy duty off-road vehicle. Moreover, a vision-based recognition algorithm of unstructured roads for this automated wheeled heavy duty off-road vehicle is presented to extract the complete feasible road area. The road recognition method consists of the multilayer perceptron self-supervised on-line learning and post-process based on a road model assumption. The experimental results show the effectiveness of the proposed system.