基于MATLAB的遥感图像棚户区识别与提取

吴敏

基于MATLAB的遥感图像棚户区识别与提取

引用本文：吴敏.基于MATLAB的遥感图像棚户区识别与提取[J].计算技术与自动化,2012,(1):97-102

摘要点击次数: 1758

全文下载次数: 125

作者	单位
吴敏	(河海大学机电工程学院，江苏常州213022）

中文摘要:以QuickBird遥感影像为数据源，研究在MATLAB环境下，利用灰度共生矩阵（GLCM）对QuickBird遥感图像棚户区纹理特征进行识别和提取的方法。首先，利用所选择的四个纹理特征统计量（对比度、能量、同质性和相关性）构建一个特征空间。然后运用非监督分类的方法（ISODATA算法）将研究区分成确定的类别数目。最后，根据实际情况，利用数学形态法对分类结果进行精确定位，从而提取得到研究区域的棚户区信息。

中文关键词:Matlab 棚户区识别灰度共生矩阵纹理特征

The Identification and Extraction of Informal Settlement Areas From Remote Sensing Images Based on MATLAB

Abstract:This paper studies the identification and extraction informal settlement areas from remote sensing images using ray level co-occurrence matrix (GLCM), realized the classification and analysis for QuickBird remote sensing image using the filter of the un-supervised classification methods under MATLAB environment. Firstly, by using the selected four texture statistics (contrast, energy, homogeneity and relevance) to construct a feature space. Secondly, the unsupervised classification method (ISODATA algorithm) will research into determining the number of classes. Finally, according to the actual situation, using mathematical morphological method for the classification results for accurate positioning, so as to extract obtained from the study area shantytown information.

keywords:Matlab mecognition of informal settlement areas gray ievel co-occurrence matrix texture feature

查看全文 查看/发表评论 下载pdf阅读器