基于激光测距雷达和机器视觉的障碍物检测  

Obstacles Detection Based on Laser Ranging Radar and Machine Vision

作　　者：王葵[1,2,3] 徐照胜[1,2,3] 颜普[1,3] 王道斌 

机构地区：[1]安徽大学电子信息工程学院,安徽合肥230039 [2]中科院合肥智能机械研究所,安徽合肥230027 [3]安徽大学计算智能与信号处理教育部重点实验室,安徽合肥230039

出　　处：《仪表技术》2012年第8期25-28,共4页Instrumentation Technology

基　　金：国家自然科学基金资助项目(61172127);高等学校博士点科研基金(20113401110006);青年科学基金项目(E080702);安徽大学211工程学术创新团队基金资助项目(KJTD007A)

摘　　要：障碍物检测问题一直是移动机器人的研究热点之一,也是机器人实现安全可靠导航的前提。文章利用激光测距雷达和机器视觉对障碍物进行联合检测,首先对摄像机标定,然后进行摄像机和激光测距雷达之间的联合标定,最后通过激光点数据将检测图像划分为几个小区域并进行图像处理。实验结果表明该方法能够有效检测出障碍物的大小信息,将检测图像划分为小区域也提高了图像处理的运算速度。Dynamic obstacle detection has been the research focus of the mobile robot, and is a prerequisite to achieve safe and reliable navigation for the robot. Joint detection of the obstacle using laser ranging radar and machine vision is introduced in this article. Camera is calibrated first, and then joint calibration between the camera and laser ranging radar is done, finally detected image is divided into several small regions through the laser point and image pro- cessing. Experimental results show that this method can effectively detect the size of the obstacle. The detected image divided into a small area also increased the speed of image processing operations.

关 键 词：障碍物检测 激光测距雷达 联合标定 激光点 图像处理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]