Abstract: The integration of remote sensing (RS) with geographic information system(GIS) is a hotspot in the field of geographic information science.A good database structure is important to the integration of RS and GIS.It would help the complete integration of RS with GIS, and also help the knowledge discovery and exploitation.It would be able to deal with the disagreement between the resolution of remotely sensed images and the precision of GIS data.In this paper, a new database structure storing the spatial data based on semantic network is presented.In this database structure, the spatial data are stored as raster data with space index, so the image processing can be done on the GIS data directly.The data are stored hierarchically according to the precision of distinguishing.Simple objects are aggregated into complex objects.The index tree depicts the relationship of aggregation, and the index pictures using 2D strings depict the topology structure of the objects.The concepts of surrounding and region are expressed clearly, so the semantic content of the landscape can be depicted well.All the factors that affect the recognition of the objects are depicted in the factor space.The factor space provides a uniform knowledge node and index node are integrated into one node.This feature enhances the ability of system in knowledge expressing, intelligent inference and association.The database structure benefits using the information of GIS for the interpretation of remotely sensed image.