Traversable terrain classification for outdoor autonomous robots using single 2D laser scans

Abstract

Interpreting laser data to allow autonomous robot navigation on paved as well as dirt roads using a fixed angle 2D laser scanner is a daunting task. This paper introduces an algorithm for terrain classification that fuses seven distinctly different classifiers: raw height, roughness, step size, curvature, slope, width and invalid data. These are then used to extract road borders, traversable terrain and identify obstacles. Experimental results are shown and discussed. The results were obtained using a DTU developed mobile robot, and the autonomous tests were conducted in a national park environment.