| 摘要 |
A system for use in a vehicle (10) for determining an indication of the type of terrain in the vicinity of the vehicle. The system comprises; means for receiving sensor output data from sensor(s) on the vehicle; means for determining a plurality of parameters in dependence on the sensor output data; a neural network algorithm (classifier) for receiving the plurality of parameters; and means for executing the neural network algorithm to provide a plurality of outputs corresponding to a plurality of different terrain types, the neural network being further configured to associate the plurality of parameters with one of the plurality of outputs, so as to determine an indication of the terrain type. The system could display the classification for viewing by a driver, and could further accept input from the driver which could be used to train the classifier in a self-learning process. The sensors could be acoustic (12,14,16) or radar (22) sensors. Independent claims are also included for calibrating (86) the classifier, in terms of weights applied to nodes, and storing the results (92) as part of the algorithm. |
