The classification of the terrain by a hexapod robot
[ 1 ] Instytut Automatyki i Inżynierii Informatycznej, Wydział Elektryczny, Politechnika Poznańska | [ P ] employee
2013
paper
english
EN This paper presents a new approach to the terrain classification by a hexapod robot using the tactile information. The data was acquired using the force/torque sensor mounted on the walking robot foot. Two types of classifiers were used and compared: the Normal Bayes Classifier (NBC) and the Classification And Regression Tree (CART). The article comprises the description of the experimental setup followed by the presentation of feature selection process and the comparison of the two classifiers’ accuracy. The classification system presented in the article allows the walking robot to recognize the type of the terrain on which it is currently walking on with over 90% accuracy.
825 - 833
WoS (15)