Path selection based on local terrain feature for unmanned ground vehicle in unknown rough terrain environment

Abstract

In this paper, we propose an autonomous navigation for Unmanned Ground Vehicles (UGVs) by a path selection method based on local features of terrain in unknown outdoor rough terrain environment. The correlation between a local terrain feature obtained from a path and a value of the path obtained from the global path planning is extracted in advance. When UGV comes to a branch while it is running, the value of path is estimated using the correlation with local terrain feature. Thereby, UGV navigation is performed by path selection under the criterion of shortest time in unknown outdoor environment. We experimented on a simulator and confirmed that the proposed method can select more effective paths in comparison with a simple path selection method.