Real time 3D terrain elevation mapping using ants Optimization Algorithm and stereo vision

Abstract

Reliable perception of terrain slope and terrain traversability is a key-feature for any off-road unmanned ground vehicle, as well as for any Driver Assistance Systems designed to work in extreme envirnoments, like mining. In this paper we want to present an innovative technique to build a 3D elevation map of the traversable terrain from a world's 3D dense data set, in real time. The 3D points are grouped into lateral and longitudinal equally spaced slices, then they are projected onto the corresponding slices' reference planes. The projections are then analyzed by a biologically inspired Optimization Algorithm able to segment points into terrains inlier and outlier; the resulting 2D terrain slopes represent an optimal terrain approximation along each slice. Finally, the 2D approximations are merged together, to create the overall 3D terrain surface.The algorithm has been successfully tested with 3D data provided by a stereo camera system mounted on a Cat R wheel loader operating in a mining environment.

Traversability analysis using terrain mapping and online-trained Terrain type classifier

“…In the literature this problem has been faced with a variety of methods, from 3D terrain analysis [1] based on stereo data to the use of some a priori knowledge of the roads visual characteristics [2]. However, none of them deals with mutable path visual characteristics that actually represent the most common scenario for a AGV moving on a long haul; the data input may be noisy and eventually new for classifiers, leading to possible misdetections.…”

A Block-based 3D Map: a Crude Sketch of Steric Environment

“…For flexible services in intelligent environments, 3D maps with value added data would be desirable. Many traditional researches on 3D maps, however, are not appropriate for real-time services since images consist of large amount of data and takes time to be manipulated [6,8].…”

Terrain mapping for off-road Autonomous Ground Vehicles using rational B-Spline surfaces and stereo vision

“…In our previous work [8] we presented a geometry based technique able to segment 3D data points into terrain inliers and outliers, without any constrains on vehicle pose and surface roughness. The segmentation is made on the basis of traversable terrain concept: any surface where the vehicle can drive on.…”

“…In [8] Ants Colony Optimizations (ACO) was used to fit points into a certain number of longitudinal and lateral 2D models, then fused together to obtain a 3D Cartesian model of the terrain. ACO works very satisfactorily, but it is computationally demanding (as well as RANSAC based techniques), since it requires many iterations for each model to reach optimal results.…”