Content area

Full text

Yibing Zhao [1] and Jining Li [2] and Linhui Li [1] and Mingheng Zhang [1] and Lie Guo [1]

Academic Editor: Valentina E. Balas

, 1, , State Key Laboratory of Structural Analysis for Industrial Equipment, , School of Automotive Engineering, , Dalian University of Technology, , Liaoning 116024, , China, , dlut.edu.cn,

, 2, , Digital Arts Department, , Dalian Neusoft Institute of Information, , Liaoning 116023, , China, , neusoft.edu.cn,

This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. Introduction

Due to the special value of military, the developed countries have researched autonomous driving car since the 1970s. At present, the United States, Germany, and Italy stand for the forefront of feasibility and practical application aspects. In 2000, the plan named Demo III of USA mainly developed the technology which was based on stereovision environment perception. Thus no matter in the day or night the Unmanned Ground Vehicles can achieve higher speed in the complex cross-country environment [1]. The key technology of this research program focuses on how to improve the ability of obstacle’s classification, for instance, how to classify bushes, rocks, grass, and water region for UGV’s automatic identification, which is playing significantly important role in the development of military utility in the future. As a matter of fact, the speed of autonomous vehicles is seriously restricted, since three-dimensional point cloud data acquisition and features matching are fairly sluggish based on stereovision. In 2007, the United States DARPA held a named “urban challenge” grand prix in Los Angeles. All autonomous driving vehicles were required to avoid obstacles, obey the traffic signal, and integrate the general traffic in the urban environment. Also, autonomous driving vehicles can automatically avoid other vehicles in this journey of 96 km—long urban model. Finally six vehicles arrived at the end point [2]. In 2009 the United States Jet Propulsion Laboratory announced obstacle detection including trunks, branches, excessive slope, negative obstacle, and water region only based on stereovision. According to each stereovision image it established a simple map, and the tag unit of the map showed the non traversable area where obstacles exist. Also the terrain height, terrain...

         

   

Show less