摘要
随着计算机视觉技术的发展,基于单目视觉的同步定位与地图创建(monocular SLAM)逐渐成为计算机视觉领域的热点问题之一。介绍了单目视觉SLAM方法的分类,从视觉特征检测与匹配、数据关联的优化、特征点深度的获取、地图的尺度控制几个方面阐述了单目视觉SLAM研究的发展现状。最后,介绍了常见的单目视觉与其他传感器结合的SLAM方法,并探讨了单目视觉SLAM未来的研究方向。
With the development of computer vision technology,monocular simultaneous localization and mapping（ monocular SLAM） has gradually become one of the hot issues in the field of computer vision. This paper introduces the monocular vision SLAM classification that relates to the present status of research in monocular SLAM methods from several aspects,including visual feature detection and matching,optimization of data association,depth acquisition of feature points,and map scale control. Monocular SLAM methods combining with other sensors are reviewed and significant issues needing further study are discussed.
出处
《智能系统学报》
CSCD
北大核心
2015年第4期499-507,共9页
CAAI Transactions on Intelligent Systems
基金
国家863计划资助项目(2006AA04Z247)
国家自然科学基金资助项目(60875050
60675025
61340046)
深圳市科技计划项目及基础研究计划资助项目(201005280682A
JCYJ20120614152234873)
教育部高等学校博士学科点专项科研基金资助项目(20130001110011)
关键词
单目视觉
同步定位与地图创建
扩展卡尔曼滤波器
计算机视觉
特征检测与匹配
monocular vision
simultaneous localization and mapping
extended Kalman filter
computer vision
feature detection and matching
