中文摘要：

如何对多源、多尺度和不同分辨率的海量影像数据进行高效存储与管理已成为当前地理信息数据库建库的关键。镶嵌数据集是当前管理栅格影像数据的新技术。本文详细阐述了镶嵌数据集管理海量数据的处理方法与入库技术流程。同时以青海湖流域为研究区，选择最新的Landsat8遥感影像数据与AsterDEM两种数据格式为例，应用新的镶嵌数据集数据处理与入库技术，实现了不同格式影像数据的镶嵌处理、数据系列产品的生成以及不同格式栅格数据的标准统一与数据建库工作。实验结果表明，使用镶嵌数据集可以高效地处理不同来源、不同时期以及不同格式的影像数据．可以实现多种类型影像数据及相关数据产品的实时构建、快速查询以及大容量数据的高效存储．说明应用镶嵌数据集在多源数据集成管理、数据显示以及网络并发访问等方面具有巨大的优势。

英文摘要：

With the development of ＂3S＂ （GPS, RS, GIS） technology, image data have been experiencing a substantial growth. It has been the linchpin of building GIS database to integrate massive image data efficiently from multi-source, multi-scale and different resolutions. This study introduced a new mosaic dataset technique for managing raster image, and also elaborated methods and technical processes of how to process the massive image data by using mosaic dataset model. Taking Qinghai Lake Basin as a case study, this paper selected two typical raster image formats including the latest remote sensing data of Landsat-8 and Aster DEM data as an example. Then, applying the mosaic dataset techniques to process and produce the series thematic dataset, the integrated image database was built for multi-source image data storage and management. The results showed clearly that using mosaic dataset can manage multi-source, different-time and different-format image data efficiently and produce various types of image products instantaneously. Consequently it has obvious advantages in integrated management for multi-source data visualization, high-speed inquiry, and concurrent access on enormous database on the internet.