Spatial knowledge databases as applied to the detection of changes in urban land use

Changes in land can involve pressing and complex problems that urban areas need to resolve. Construction of a land use or land cover mechanism is an essential way of dealing with those problems. Based upon the difficulties that occur in investigations of land use, this study will focus on the development of a spatial knowledge database. The major function of a spatial knowledge database is to integrate spectral, GIS, aerial photos, and DEM information. In this study data from a series of periods for SPOT‐HRV XS images (1993, 1996, 1998) were collected. A decision‐tree concept was built into the image‐classification database to resolve the uncertainty of images in urban areas. The aim of this study is to provide an objective solution for efficiently classifying images in urban areas.