We present an approach to road recognition based on Kalman filtering and EM
theory resulting in a fast recursive filter that can adaptively track unpa
ved roads. The road model uses linear dynamic equations and an ANN front en
d detects road boundaries. Kalman filters construct search neighborhoods an
d adapt the ANN to road conditions. The road model adapts to motion dynamic
s using EM. Experimental results are presented. (C) 2002 Elsevier Science B
.V. All rights reserved.