CASIA OpenIR  > 毕业生  > 博士学位论文
虚拟现实中的跟踪技术研究
其他题名Research on the Tracking Technology for Virtual Reality
严志刚
学位类型工学博士
导师原魁
2005-05-26
学位授予单位中国科学院研究生院
学位授予地点中国科学院自动化研究所
学位专业控制理论与控制工程
关键词虚拟现实 跟踪技术 自主跟踪技术 电磁跟踪技术 组合跟踪技术 Virtual Reality Tracking Technology Self-tracking Technology Electromagnetic Tracking Technology Hybrid Tracking Technology
摘要虚拟现实中的跟踪技术是虚拟现实技术和增强现实技术的重要组成部分, 对虚拟现实系统和增强现实系统的实现、完善以及发展具有重要意义。本文在国家自然科学基金“虚拟外科手术系统中的人机接触交互技术”课题资助下,对虚拟现实中常用的跟踪技术进行了研究。在这些技术中,自主式跟踪技术、电磁跟踪技术和组合跟踪技术是目前主要的跟踪方式,他们获得了广泛应用并具有广阔的发展前景。本文重点对这三种技术进行了研究,具体研究内容主要包括以下几个方面: 第一.论文首先回顾了虚拟现实技术和虚拟现实中跟踪技术的发展历程;另外,对目前虚拟现实中常用跟踪技术的技术原理和发展概况进行了综述;最后,分析比较了各种技术的优缺点,并讨论了虚拟现实中跟踪技术的发展方向。 第二.对自主式跟踪技术进行了研究,提出了一种改进的基于地磁场和重力场的姿态跟踪算法。该算法减少了传统算法中所需要的信息量,简化了系统设计。经过实验验证,证明该方法在减少了信息量的情况下,同传统算法相比有相同的姿态跟踪性能。 第三.由于电磁跟踪模型在电磁跟踪研究中占有重要地位,本文对常用的电磁跟踪模型进行了研究,并提出了一种改进的跟踪模型。通过仿真分析,比较了两种电磁跟踪模型的差异,并通过实验验证了这种模型的有效性。 第四.对电磁跟踪算法进行了研究,提出了一种基于分布式参数的电磁跟踪算法。同传统的将系统各参数视为一致的跟踪算法相比,这种算法更符合实际跟踪系统的物理本质。同时,该算法也简化了系统硬件设计。为了准确获得这些参数,本文也提出了一种新的获得电磁跟踪系统分布式参数的简单方法。本文将改进的电磁跟踪模型和这两种算法相结合应用于Fastrak系统和自开发系统。实验数据证明,本文所提算法和改进模型是有效的,并且提高了系统跟踪性能。 第五.对组合式跟踪技术进行了研究。总结了组合式跟踪系统常用的组合方式、数据融合技术和运动模型,在此基础上,提出了一种新的基于地磁、惯性和单目视觉的组合跟踪方法。同其它方法相比,该方法降低了系统复杂度。实验验证表明,该方法提高了系统鲁棒性,可以很好的跟踪物体的6DOF运动信息。 最后,对本文所做的研究工作及取得的研究成果进行了总结,指出了需要继续开展的工作。
其他摘要Tracking technology for virtual reality plays an essential role in virtual reality technology and augmented reality technology. It has significant impact on the realization, perfection and the future development of virtual reality and augmented reality. Sponsored by the National Natural Science Foundation of China,this dissertation makes a study on the available tracking technologies for virtual reality. Among these technologies, the self-tracking techonolgy, electromagnetic tracking technology and hybrid tracking technology are the major technologies . So this dissertation devotes research on these three technologies. The main contents of this dissertation are listed as follows: Firstly, the dissertation briefly reviews the developing history of the virtual reality technology and the tracking technology. Then the principle and the developing status of the available tracking technology are surveyed. The virtues and the shortcomings of each technology are also analyzed. Finally, the direction of future development of tracking technology is discussed. Secondly, the dissertation studied the self-tracking technology. An improved orientation tracking algorithm based on the earth magnetic field and the gravity field is proposed. Compared with the traditional algorithm, the algorithm needs less information, which can simplify the system design. The experiment results show that the improved algorithm has same performance as the traditional algorithm . Thirdly, this dissertation makes a study on the commonly used model for electromagnetic tracking. An improved electromagnetic tracking model is proposed. The differences between these two models are presented by simulation. Some experiments are also carried out and the results indicate the improved model is valid. Fourthly, the electromagnetic tracking algorithms are studied. An improved electromagnetic tracking algorithm is proposed. Compared with the traditional algorithm which takes the system parameters as same, it is more accordant to the nature of the tracking system. In order to get the system scattering parameters precisely, a simple method to get these parameters is also proposed. In the dissertation, these two algorithms are combined with the improved model and are applied to the Fastrak system and the tracking system which was developed by us. The experiment results indicate the two algorithms and the improved model are valid and can improve the performance of the tracking system. Fifthly, the hybrid tracking technology is studied. The combination mode, data fusion technology and dynamic model are surveyed. Based on this, a novel hybrid tracking algorithm is proposed. Compared with other hybrid tracking algorithms, the algorithm makes the system simpler. The experiment results indicate that the algorithm can improve the robustness of the system. Finally, the achieved research results are summarized and the future work is addressed.
馆藏号XWLW971
其他标识符200418014690006
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/5870
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
严志刚. 虚拟现实中的跟踪技术研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2005.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
CASIA_20041801469000(1257KB) 暂不开放CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[严志刚]的文章
百度学术
百度学术中相似的文章
[严志刚]的文章
必应学术
必应学术中相似的文章
[严志刚]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。