文章基本信息

标题：A Knowledge-Based System for Unsupervised Classification of High Resolution Multispectral Satellite Images
本地全文：下载
作者：Andrea Baraldi
期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：1992
卷号：XXIX Part B2
页码：59-66
出版社：Copernicus Publications
摘要：A knowledge-based, hierarchical, unsupervised classification scheme forhigh resolution multispectral satellite images is proposed. This scheme,which finds its conceptual bases in the work of Nagao and Matsuyama(1980) for structural analysis of aerial photographs, introduces a newfiltering algorithm which is able to preserve fine linear structures of theimage.The classification products are: 1) a raster image where detectedoutput classes are characterized by some specific information contentand/or by spectral separability (by cluster analysis); 2) property tablesdescribing each elementary region in terms of descriptive and geometricattributes; 3) geometry of output elementary regions converted tovector format.An example of the application of this classification scheme to aLandsat TM multispectral image is presented.Key Words: satellite image classification, knowledge-based classification,image understanding.