影像数据具有数据量大、种类繁多、结构复杂、生产速度快等特点，给数据的管理和分发工作带来挑战。目前很多单位使用基于文件的方式管理影像数据，当数据量超过一定规模后这种方式就暴露出一系列问题，严重影响到影像的精确检索、高效分发和利用，降低了影像的利用率和管理效率。本论文在此背景下研究探讨基于矢量GIS技术利用关系数据库并结合FTP文件传输协议来进行影像数据的管理。论文主要介绍影像管理的研究现状及其发展，讨论海量影像管理的历史与现状，概述了海量影像管理相关的分布式存储、网络传输、图形绘制等一系列理论。着重讨论影像数据库在建立过程中的各项关键技术：影像数据元数据管理；多来源影像的坐标系统一致性问题；多分辨率的影像金字塔结构；基于矢量GIS系统叠加的各种空间查询的设计；影像数据快速显示技术等。在实践环节，依托国家西部测图工程管理平台项目中多源存档影像资料管理的实际需求进行调研、分析并具体设计开发了一个可用于实际生产管理的存档影像管理的软件系统。该系统的开发、测试和运行实践进一步论证了本文提出的影像管理模式的可行性和合理性。论文的最后对本文所作的研究工作进行了总结，并对进一步的研究内容进行了展望。

Image data with a large quantity of data, a wide range of complex structures, and faster production features bring challenges to the management and distribution work. At present, many organizations use file-based system management the image data, when the volume of data over a certain size this way have exposed a series of problems that seriously affects the image precision retrieval, efficient distribution and use, reducing the utilization and efficiency of management. Under this background, this paper is focused on the study of using relational database technology combined with FTP file transfer protocol based on vector GIS for image data management. This thesis mainly introduced the image management-related research and development status, discussed the history and current situation of the massive image management and outlined distributed storage, network transmission, graphics rendering and a series of theories related to the huge image management. Focused on the key technologies in the process of the images database establishing: image data metadata management; multi-source image consistency of the coordinate system; multi-resolution image pyramid structure; the image space inquiry design based on the superposition of vector GIS system; image data fast display technologies. In the practice stage, based on the State's Western Mapping Project Management Platform’s image data management needs, doing the investigation and actual research, design and development a specific operational archive image management software system and deploy the trial version of this system. After the development and testing practice for the systems, it further demonstrated that the image management model presented by this paper is feasible and rationality. In end of the paper, concluded the paper's research, and discussed the contents of the future.