However the technique is always inefficient when visibility is interfered due to climatic conditions. The reaction based classification is dependent on the sensor measurement like wheel sinkage, vehicle vibrations, and wheel slip (Garratt and Chahl, 2011). However, some of the interactions of vehicle terrain like the wheel sinkage and slip is not easy to measure accurately. Therefore, the reaction based terrain in normally performed with the help vehicle vibrations. This is because they are easy to measure with the help of accelerometers and inertial sensors. An unmanned vehicle ought to have a complete understanding of the features and the terrain around it in order to be able to navigate multifarious environmental safety (Brooks and Iagnemma, 2010). In other cases, the terrain can only be understood when speed and optimality is sacrificed. This chapter is a literature review that will utilize the measured acceleration values in the vertical direction of the vehicles to classify the terrain traversed. The literature review will be divided into a number of parts discussing wheel sinkage methods, vision based methods and different vibration methods aimed at classifying the data.
Terrain classification for ground vehicles are normally categorized as reaction based methods, vision based method, and combination of a reaction and vision based methods. The reaction and vision based approaches are quite analogous to the recognition of human drivers of a terrain in respect to what can be seen and felt through the reaction of the vehicle during the terrain traversal. Vision-based terrain is performed with the help of laser range finders and cameras.