首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种基于TM影像的湿地信息提取方法及其变化检测
引用本文:李芳芳,贾永红.一种基于TM影像的湿地信息提取方法及其变化检测[J].测绘科学,2008,33(2):147-149.
作者姓名:李芳芳  贾永红
作者单位:1. 武汉大学遥感信息工程学院,武汉,430072
2. 武汉大学遥感信息工程学院,武汉,430072;武汉大学测绘遥感信息工程国家重点实验室,武汉,430072
摘    要:近年来,有着"地球之肾"之称的湿地,其对于生态环境的特殊性和重要性受到全世界的极大关注,湿地研究己成为当前国际上的热门领域。本文结合缨帽变换及决策树分类法提出了一种新的湿地提取方法。首先对原始影像进行缨帽变换,提取其湿度分量,利用湿度分量进行分级提取,区分湿地与陆地。接下来在深入分析研究区影像光谱特征规律的基础上,建立了湿地信息提取决策树模型,用于区分不同类型湿地。最后,对两个时相的湿地分布影像进行了变化检测,提取出了湿地的变化范围。

关 键 词:湿地  K-T变换  决策树  变化检测
文章编号:1009-2307(2008)02-0147-03
收稿时间:2006-12-20
修稿时间:2006年12月20

Study on method of extracting wetland and its changed area based on Landsat TM images
LI Fang-fang,JIA Yong-hong.Study on method of extracting wetland and its changed area based on Landsat TM images[J].Science of Surveying and Mapping,2008,33(2):147-149.
Authors:LI Fang-fang  JIA Yong-hong
Abstract:Wetlands, which are named as "kidney of the earth", have caught human’s eyes for its unique and important functions for ecological environment. At present, wetland research has been a very important research field in remote sensing. In this paper, we propose a wetland extraction algorithm that combines Tasseled Ca P(TC) transformation and decision tree classification (DTC). The humidity component comes from TC transformation of multispectral images is used to differentiate wetlands from land. After deeply analyzing wetlands’ characteristics and their spectral features, we develo Pa DTC model to distinguish different wetland types. At last, to extract changed wetland areas, we present a change detection method based on two phases of the same area’s wetland classification images.
Keywords:wetland  K-T transform  DTC  change detection
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号