Road and off road terrain classification for autonomous ground vehicle

Abstract

The ongoing development of Autonomous Ground Vehicle technologies necessitates for classification of terrain as road and off road to identify the drivable path and optimal velocity for traversal of the vehicle. Terrain consists of different texture types, classification of terrain into different classes is a difficult and challenging task. In this paper the feature set extraction and classification approaches are explored, and a novel vision based method for classification of terrain as two classes {road and off road} is presented. The proposed feature extraction process utilizes both color and texture distributions and is combined with a trained multi-layer feed forward neural network based supervised classifier to categorize the terrain. The algorithm is trained and tested on video data obtained from front looking cameras mounted on a vehicle and it is observed that an optimal performance of 93% correct classification is achieved using the proposed methods.