Extracting drivable surfaces in outdoor 6D slam

Abstract

A 6D SLAM method for extracting drivable surfaces in an outdoor environment, is discussed. In an outdoor terrain, actively looking for navigable surfaces is mandatory, as ragged ground, potholes, or steps are to be expected at any place. An algorithm is designed for interpreting 3D scans and labeling floor points by computing the gradient between a point in a given cylindrical coordinate system. This labeling has proved to be robust against uneven and non-horizontal ground because of wide sets of drivable points with only small height differences. 3D geometry maps of outdoor terrain are now able to mark paths and other geometrically drivable surfaces in the maps.