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：2002
卷号：XXXIV Part 3 B
页码：116-119
出版社：Copernicus Publications
摘要：Using a set of colour-infrared aerial photos, we compare a newly developed neural net based clustering method with a method based on the classical ISODATA algorithm. The primary focus is on the detection of buildings and it shows that while the traditional method has an advantage in splitting background from foreground, the neural net based method results in a much more uniform segmentation of the input images
