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

基于 C4.5决策树的多特征遥感分类方法
引用本文:曹洪涛,张拯宁,李明,李器宇,陈浩.基于 C4.5决策树的多特征遥感分类方法[J].测绘工程,2016,25(3):73-76.
作者姓名:曹洪涛  张拯宁  李明  李器宇  陈浩
作者单位:天津航天中为数据系统科技有限公司,天津,300301;天津航天中为数据系统科技有限公司,天津,300301;天津航天中为数据系统科技有限公司,天津,300301;天津航天中为数据系统科技有限公司,天津,300301;天津航天中为数据系统科技有限公司,天津,300301
摘    要:以钱塘江流域为研究区域,利用2010年ETM,MODIS和DEM多源数据,进行土地利用分类研究。在分析土地类型的光谱特性和植被指数年度变化基础上,运用光谱指数法和代数法从数据中提取各种土地覆被类型特征。利用WEKA软件平台下的C4.5决策树算法构建决策树分类模型,对钱塘江流域土地覆被类型进行分类研究,取得较高的分类精度。

关 键 词:多特征  C4.  5决策树  遥感影像  WEKA

Remote sensing classification with multi-feature based on C4. 5 decision tree method
CAO Hongtao,ZHANG Zhengning,LI Ming,LI Qiyu,CHEN Hao.Remote sensing classification with multi-feature based on C4. 5 decision tree method[J].Engineering of Surveying and Mapping,2016,25(3):73-76.
Authors:CAO Hongtao  ZHANG Zhengning  LI Ming  LI Qiyu  CHEN Hao
Abstract:Taking Qiantangjiang Basin as the study area ,land‐cover classification reseach is conducted in this paper using Landsat ETM ,MODIS and DEM .Based on analysing spectral characteristics and annual changes in vegetation index ,the land‐cover classification of Qiantangjiang Basin has been done .Based on C4.5 decision tree method from software WEKA the land‐cover type features from data are analyzed with the spectral index method and the algebraic method .Compared with the maximum likehood classification and Neural net classification ,the results show that classification accuracy is better .
Keywords:muti-feature  C4  5 decision tree classification  remote sensing images  WEKA
本文献已被 CNKI 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