SIA OpenIR  > 其他
Alternative TitleResearch on Indoor Service Robot Localization Based on Binocular Stereo Vision
Thesis Advisor曲道奎
Keyword服务机器人 双目立体视觉 摄像机标定 立体匹配 机器人定位
Call NumberTP242.3/Z33/2009
Degree Discipline模式识别与智能系统
Degree Name博士
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract智能服务机器人是我国将大力发展的四大先进制造技术之一。视觉传感器的不确定性给机器人的实时环境感知和定位问题带来了巨大的挑战,直接影响到机器人导航的效果。所以探索服务机器人立体视觉定位技术显得非常必要。本文的研究属于863重点项目“机器人模块化体系结构设计”。 1.在立体相机的标定过程中,提出了一种节点提取方法,将节点提取的精度提高到了亚像素级别。 2.本文在已有图割算法的基础上提出了一种改进匹配算法,利用局部匹配算法得出的像素差作为先验值,去除大量不相关节点和边缘,缩减匹配所用的时间。 3.在基于路标的机器人自定位系统中,设计了三种用于机器人定位导航的路标,提出了一种组合算法识别路标;推导了单路标和多路标情况下移动机器人定位的全局坐标公式;实验验证了在多路标的情况下,机器人自定位的精度有显著提高。 4.多传感器融合定位系统中主要论述了融合视觉传感器和里程计传感器的移动机器人定位算法。提出了一种改进粒子的重采样过程,引入了过收敛验证和均匀性验证用于检验采样分布与感知信息的匹配程度,引入不同的重采样方法。实验分析表明了改进定位方法能够有效地提高机器人的全局定位精度和解决机器人绑架问题。
Other AbstractIntelligent service robot is one of four major advanced manufacturing technologies which China will greatly develop in the following years. The character of vision system is that it can capture a great deal of information of the environment, but the uncertainty of the vision system has brought the enormous challenge to the robot navigation, which influence the autonomous navigation directly, therefore it is very difficult to find out a robust navigation algorithm to be used in the indoor environment under various situations. So to explore the home service robot localization based on stereo vision in family's dynamic environment, which is very essential at present. The research is funded by the National High-tech Research and Development Program (863 Program) "module structural design of the robot system ". 1. In the camera calibration, we have proposed an improved method to extract the nodes which utilizes Harris node extraction algorithm and curve fitting algorithm. It is proved by experiments that the improved algorithm can achieve sub- pixel precision. 2. Stereo matching is the most difficult part in the stereoscopic vision. In this paper we have proposed an improved stereo matching algorithm based on graph cut. The algorithm reduced much unrelated nodes and edges by a local matching met empirical value efficiently and saved much stereo matching time. 3. In the mobile robot navigation system, we have designed three kinds of landmarks and proposed the associated recognition algorithm which is tested to recognize the landmarks in most cases. The robot localization formula by one landmark and multiple landmarks are deduced in this paper, It has been proved that the locating precision has been improved by multiple landmarks than single landmark. 4. The integrated robot localization system mainly investigates the locating algorithm based on stereo vision system and dead reckons system. We propose an improvement particle resample course, introduce two courses to verify the particles. The experiment analysis indicates that improving Monte Carlo can improve the precision of the robot global localization and solve robot kidnapping problem, meanwhile which can improve the robot independence and locating efficiency, and has a good calculation performance.
Contribution Rank1
Document Type学位论文
Recommended Citation
GB/T 7714
张令涛. 室内服务机器人基于双目立体视觉系统的定位技术研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2009.
Files in This Item:
File Name/Size DocType Version Access License
室内服务机器人基于双目立体视觉系统的定位(1664KB) 开放获取LicenseApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[张令涛]'s Articles
Baidu academic
Similar articles in Baidu academic
[张令涛]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[张令涛]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.