基于单目视觉直线跟踪的SLAM实现

作者：王隽永,何 衍，陈家乾 日期：2007-07-09/span> 浏览：3481 查看PDF文档

                               基于单目视觉直线跟踪的SLAM实现
                                       王隽永,何 衍，陈家乾
                    （浙江大学 电气工程学院，浙江 杭州 310027）

摘 要：同时定位和地图生成问题（SLAM）是移动机器人的一个基本问题，有着广阔的应用背景。由于视觉传感器具有成本低、信息丰富的特点，是当前SLAM问题研究的一个热点。针对视觉传感器视野小、远距离观测精度低的问题，提出了一种基于直线跟踪的单目视觉SLAM算法。该算法通过对直线特征的近距离观测和跟踪，保证了定位的精度和地图生成的效率。在Pioneer3dx机器人平台上进行的实验研究表明，该算法尤其适合于结构化环境几何地图的构建。
关键词：同时定位和地图生成；扩展卡尔曼滤波；移动机器人；直线特征

          Realization of SLAM based on monocular vision and lines tracking
                             WANG Juanyong, HE Yan, CHEN Jiaqian
(College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China)

Abstract: Simultaneous localization and mapping (SLAM) is a basic problem of mobile robot. It could be applied in many areas. As the camera is lowcost and rich in information, it becomes one of the most popular issues of SLAM. According to the smallvision of the camera and low accuracy of long distance observation, the SLAM  algorithm using monocular vision and lines tracking was presented. Through closely linear observation and linefeature tracking, the accuracy and efficiency were improved. Experiment results and analysis on the Pioneer3dx mobile robot platform show that, the algorithm is suitable to mapping in structure environment.
Key words: simultaneous localization and mapping (SLAM); extended kalmanfilter (EKF); mobile robot; linefeature

参考文献(Reference)：
［1］LI M H, HONG B R, LUO R H, et al. A novel method for mobile robot simultaneous localization and mapping［J］.浙江大学学报A:英文版,2006,7(6):937-944.
［2］庄 严,王 伟,王 珂,等.移动机器人基于激光测距和单目视觉的室内同时定位和地图构建［J］.自动化学报,2005,31(6):925-933.
ZHUANG Yan, WANG Wei, WANG Ke, et al. Mobile robot indoor simultaneous localization and mapping using laster range finder and monocular vision[J]. Acta Automatica Sinica,2005,31(6):925-933.
［3］CASTELLANOS J A, NEIRA J. Multisensor fusion for simultaneous localization and map building［J］.IEEE Trans on Robotics and Automation,2001,17(6):908-914.
［4］TOMATIS N, NOURBAKHSH I, SIEGWART R. Hybrid simultaneous localization and map building: a natural integration of topological and metric［J］. Robotics and Autonomous Systems,2003,44(1):3-14.
［5］周庆瑞，原 魁，刘俊承.移动机器人视觉导航控制研究［J］.计算机工程与应用,2005,41(15):5-8.
ZHOU Qingrui, YUAN Kui, LIU Juncheng. Research on control strategy of mobile robot navigation based on vision[J]. Computer Engineering and Applications,2005,41(15):5-8.
［6］韩 锐，李文锋.一种基于线特征的SLAM算法研究［J］.自动化学报，2006,32(1):43-46.
HAN Rui, LI Wenfeng. Linefeaturebased SLAM algorithm[J]. Acta Automatica Sinica,2006,32(1):43-46.