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：2006
卷号：XXXVI Part 3
出版社：Copernicus Publications
摘要：Traditionally, an unsupervised classification divides all pixels within an image into a corresponding class pixel by pixel; the number of clusters usually needs to be fixed a priori by a human analyst. In general, the spectral properties of specific information classes change with the seasons, and therefore, the relation between object class and spectral cluster is not constant over time. In addition, relations for one image can in general not be extended to others. Thus, even if the number of clusters is correctly fixed for one image at one instance in time, the results cannot be transferred to other areas or epochs. In this study, a heuristic method based on Genetic Algorithms (GA) is adopted to automatically determine the number of cluster centroids during unsupervised classification. The optimization is based on the Davies-Bouldin Index (DBI). A software programme was developed in MATLAB, - and the GA unsupervised classifier was tested on an IKONOS satellite image. The classification results were compared to conventional ISODATA results, and to ground truth information derived from a topographic map for the estimation of classification accuracy
