Classification and Modeling of Forested Terrain from Unmanned Ground Vehicles

To operate autonomously, unmanned ground vehicles (UGVs) must be able to identify the load-bearing surface of the terrain (i.e. the ground) and obstacles. Current classification, modeling and navigation techniques work well for structured environments such as urban areas, where there are roads and obstacles that are usually predictable and well-defined. However, autonomous navigation in forested terrain presents many new challenges due to the variability and lack of structure in natural environments.

This project focuses on:

(*) Using LIDAR sensing to classify and model the ground-plane and main tree stems (i.e. trunks)
(*) Utilizing trinocular vision to enhance classification and modeling techniques
(*) Implementing a simultaneous localization and mapping (SLAM) algorithm, using the location of main tree stems as inputs
(*) Autonomous navigation of a UGV in forested terrain

Experimental testing for this project has been performed on a MobileRobots P3-AT robot platform.