Natural landmark-based autonomous vehicle navigation

Abstract

This article describes a natural landmark navigation algorithm for autonomous vehicles operating in relatively unstructured environments. The algorithm employs points of maximum curvature, extracted from laser scan data, as point landmarks in an extended Kalman filter. A curvature scale space algorithm is developed to locate points of maximum curvature. The location of these points is invariant to view-point and sensor resolution. They can therefore be used as stable and reliable landmarks in a localization algorithm. This information is then fused with odometric information to provide localization information for an outdoor vehicle. The method described is invariant to the size and orientation of the range images under consideration (with respect to rotation and translation), is robust to noise, and can reliably detect and localize naturally occurring landmarks in the operating environment. The method developed is generally applicable to a range of unstructured environments and may be used with different types of sensors. The method is demonstrated as part of a navigation system for an outdoor vehicle in an unmodified operating domain.