面向对象分类的决策树方法探讨——以Landsat-8OLI为例
发布时间：2019-06-22 11:08
【摘要】：针对目前遥感影像分类中面向对象和决策树相结合的研究较少的情况,该文提出基于C5.0决策树的面向对象分类方法,并以广州市从化区进行实证研究。基于Landsat-8OLI影像数据,采用面向对象分类对影像进行多尺度分割,提取出影像对象的光谱、纹理特征以及影像对象相对应的DEM信息;然后利用C5.0决策树根据特征信息来挖掘分类规则;最后根据规则对分割后影像进行分类。结果表明,基于C5.0决策树的面向对象影像分类精度高、效果好,总体精度和Kappa系数分别为89.75%和87.5%。该方法可准确、快速地提取土地利用/覆被信息。
[Abstract]:In view of the fact that there is little research on the combination of object-oriented and decision tree in remote sensing image classification, this paper proposes an object-oriented classification method based on c5.0 decision tree, and makes an empirical study on Guangzhou slave area. Based on the Landsat-8OLI image data, the multi-scale segmentation of the image is carried out by object-oriented classification, and the spectral and texture features of the image object and the corresponding DEM information of the image object are extracted; then the c5.0 decision tree is used to mine the classification rules according to the feature information; finally, the segmented image is classified according to the rules. The results show that the object-oriented image classification based on c5.0 decision tree has high accuracy and good effect, and the overall accuracy and Kappa coefficient are 89. 75% and 87. 5%, respectively. This method can extract land use / cover information accurately and quickly.
【作者单位】： 四川省万源市国土资源局;广州大学地理科学学院;
【基金】：国家自然科学基金项(41171070) 广州大学研究生项(2014YJS01)
【分类号】：P237
[Abstract]:In view of the fact that there is little research on the combination of object-oriented and decision tree in remote sensing image classification, this paper proposes an object-oriented classification method based on c5.0 decision tree, and makes an empirical study on Guangzhou slave area. Based on the Landsat-8OLI image data, the multi-scale segmentation of the image is carried out by object-oriented classification, and the spectral and texture features of the image object and the corresponding DEM information of the image object are extracted; then the c5.0 decision tree is used to mine the classification rules according to the feature information; finally, the segmented image is classified according to the rules. The results show that the object-oriented image classification based on c5.0 decision tree has high accuracy and good effect, and the overall accuracy and Kappa coefficient are 89. 75% and 87. 5%, respectively. This method can extract land use / cover information accurately and quickly.
【作者单位】： 四川省万源市国土资源局;广州大学地理科学学院;
【基金】：国家自然科学基金项(41171070) 广州大学研究生项(2014YJS01)
【分类号】：P237
【相似文献】
相关期刊论文 前10条
1 杜福光;;面向对象分类法在唐山南湖遥感信息提取中的优越性分析[J];唐山师范学院学报;2014年02期
2 杨叶涛;王迎迎;曾又枝;;基于面向对象影像分析的景观连接特征多尺度提取研究[J];测绘与空间地理信息;2012年12期
3 王代X;王志恒;杜霄蒙;陈雨苗;;基于面向对象技术的低山区公路信息提取研究——以广西省玉林市为例[J];天津城建大学学报;2014年02期
4 华一新，江南;面向对象GIS工具软件的开发管理[J];解放军测绘学院学报;1996年04期
5 武晓莉;高小力;;面向对象技术及其在GIS中的应用[J];测绘科技通讯;1996年04期
6 胡桑;;面向对象技术在制图中的应用[J];测绘科技通讯;1995年02期
7 王兰辉;李彬;;面向对象GIS建模的原则与方法[J];科技致富向导;2011年02期
8 蒋志华,林喜荣;面向对象技术在空间信息系统中的应用[J];电脑与信息技术;1995年04期
9 李s，
本文编号：2504522
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2504522.html
