Abstract: In order to organize and manage large volumes of remote sensing data more effectively, the authors have made an all-round study of the present style in organization and management of image data. The functional requirement was analyzed for remote sensing image background database. The database system was divided into four major functional modules, i.e., the import and upload module, the query and search module, the export and download module and the manage module. The distributed three-tier C\S structure was adopted as the system structure. The metadata information which can meet the needs of search and query was chosen. The database table was designed for the storage of data. The SuperMap SDX+ spatial database engine was employed in this study. The pyramid technology and spatial index technology were utilized, which constituted vital technologies in this study. The building of the pyramid will benefit the quick view of the image. The creation of the spatial index can help query image more efficiently. Oracle database was used as the database platform, and C#2.0 was used as the development language. The SuperMap Objects component geographic information system was utilized for the development of platform, the database was accessed by Supermap SDX+ spatial database engine which was built in SuperMap Objects, and the remote sensing image database system was developed and implemented for mineral resources. The system automatically creates the pyramid and the default space index for image data when the data are imported into database. The