【作者】 刘潺；

【导师】 陈荦；

【作者基本信息】 国防科学技术大学 ， 摄影测量与遥感， 2011， 硕士

【摘要】 21世纪以来,随着国际上又一次探月高潮的掀起,我国陆续发射了嫦娥一号与嫦娥二号探月卫星,获取并掌握了大量的月球探测空间数据。这些数据呈现出海量、多源、分布式存储等特点。因此,研究月球探测空间数据的集成管理方案,优化其检索方式,对我国探月工程的深入开展以及未来行星探测任务的规划具有重要的科研价值和现实意义。论文以月球探测空间数据集成与检索技术为主题,对以下内容进行了研究。首先,论文概述了空间数据集成的技术,提出了一个面向月球探测空间数据的集成管理框架,并根据这一框架设计了月球探测空间数据的编目模型与模式映射实现方法,详细描述了分布式环境下的月球探测空间数据集成如何实现。然后,论文分别从属性特征关联与视觉特征出发,描述了两种类型的检索。基于属性特征关联的检索介绍了属性关联的模型、检索算法以及具体实现;基于视觉特征的检索阐述了特征的提取、数据的划分以及索引的构建三个检索流程。最后,论文设计和实现了月球探测空间数据集成管理原型系统,展示了系统的功能,验证了本文关键技术用以解决月球探测空间数据集成与检索技术的可行性、可靠性和高效性。更多还原

【Abstract】 With the international prevalence of lunar exploration, our nation has successively lunched the Chang’e-1 Lunar Probe and Chang’e-2, the second one, to obtain a large amount of probed lunar spatial data which are massive, multi-source and distributive. So research on integration and management of probed lunar spatial data and optimized data index have significant realistic and scientific research value on carrying on lunar probing project and planning future planetary probing tasks. This paper focuses on the topic of integration and retrieval for lunar spatial data. And related works are specified as follows:Firstly, the integrated spatial data management technique is addressed, and then a catalogue model based on probed lunar spatial data is designed. According to this model, an integrated framework applied in managing probed lunar spatial data is proposed and the procedure of integrating distributive lunar spatial data is described.Then, two kinds of retrieval are analyzed based on attributive and visualized characters respectively. Implementation of retrieval based on attributive characters is introduced from aspects of three different data, PDS, raster-vector and media data. Image retrieval based on visualized characters is studied through the process of character extraction, data partition and indexing.At last, a prototype system is designed and implemented. Function modules are presented, which verify the feasibility, reliability and efficiency of the methods provided in this paper.更多还原