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：2005
卷号：XXXVI-8/W27
出版社：Copernicus Publications
摘要：This article introduces a new and effective method for urban zone segmentation in satellite imagery. The proposed method consists in filterin g a band of the multi-spectral image (from the SPOT satellite) with a quadrature bandpass isotropic filter. The filter bandwith and tuning frequency are determined from the source images by a simple experimental analysis. The final image can be obtained by low -pass filtering the quadrature filter's output, or by a Bayesian method using a Hidden Markov Measure Field Model. Our experiments in SPOT imagery at 10 meters resolution, report classification errors of around 5%. Sea zones containing similar features to those found in urban areas are removed by a neural network
关键词：Variance; Urban Zones; Isotropic Filters; Segmentation; Neural Network; Frequency
