Feature Extraction for Terrain Classification with Crawling Robots

Abstract

In this paper, we address the problem of terrain classification using a technically blind hexapod walking robot. The proposed approach is built on top of the ex- isting method based on analysis of the feedback from the robot's actuators and the desired trajectory. The formed method uses features for the Support Vector Machine clas- sification method that assumes a regular time-invariant gait to control the robot. However, such a gait does not allow the robot to traverse rough terrains, and therefore, it is nec- essary to consider adaptive motion gait to deal with small obstacles, which is, unfortunately, not a regular gait with some fixed predefined period. Therefore, we propose to al- ter the features extraction process to utilize the terrain clas- sification method also for an adaptive motion gait, which enables the robot to traverse rough terrains. The proposed method has been experimentally verified on several ter- rains that are not traversable by a default regular gait. The achieved results not only confirmed the high accuracy of the terrain classification as the existing approach, but also expanded the area of operation of a hexapod walking robot into more challenging terrains.