基于KDD和GIS的遥感图像专题分类方法探讨

Remote Sensing Image Thematic Classification Based on KDD and GIS

利用数据库领域的最新成果 KDD的思想和方法 ,初步探讨了从

The precision of Remote Sensing (RS) image thematic classification strongly impacts its potential power to update Geographical Information System (GIS) spatial database. Thus how to improve the thematic classification accuracy is a key study in the domain of RS & GIS. Image interpretation expert system using both the spectrum information carried by RS imagery and the knowledge of experts or the auxiliary geographical data, is an effective way to this end. But the difficulties in knowledge acquisition and the regional limitation of knowledge impede the practical use of this method. With the development of GIS, the existing data in GIS system is a new source of knowledge. Absorbed the up to date research findings on Knowledge Discovery from GIS spatial Database, this paper systematically discussed how to discover knowledge from GIS for RS thematic classification. The knowledge includes attribute dependent knowledge, spatial relations between geographical objects, shape information of these objects, knowledge about spatial and temporal change information and human influence pattern. This paper also discussed the methods for translating the knowledge to productive rules and for evaluation the rules. The system structure of such a integrated system is also discussed in general.