Natural Terrain Detection and SLAM Using LIDAR for UGV

This paper describes a natural terrain detection algorithm and a SLAM algorithm using a LIDAR sensor for an unmanned ground vehicle. We describe how features are detected from natural terrain, and then we localize the vehicle's position and compose a map with the detected features. The LIDAR equipped on the experimental vehicle to scan natural terrain. The scan data is included many kinds of intrinsic disturbance on uneven terrain: a banded tree, a branch of a tree, uniform size of bush, undefined or unexpected objects. We apply a RANSAC (RANdom SAmple Consensus) algorithm to discriminate ground point cloud data and object point cloud data, and then separate bush point cloud data and tree point cloud data by two combination algorithms; GMM (Gaussian Mixture Model) and EM (Expectation Maximization). Both GMM and EM algorithms are for extracting features and classifying groups, respectively. We propose the double FCM (Fuzzy C-mean clustering) algorithm to robustly estimate the number of trees and its position. The Extended Kalman Filter approach to simultaneous localization and mapping (EKF-SLAM) is composed of extracted tree features. The mahalanobis distance is applied to remain consistency for feature correspondence which is for data association. Finally, we show the results which is experienced in a tree-filled mountain.