|
|||||
|
|
| 基于决策二叉树的多光谱影像混合像元分类方法 | |
| 引用本文： | 李明奎,王建丽,王明孝,齐建成.基于决策二叉树的多光谱影像混合像元分类方法[J].地理空间信息,2010,8(3):100-102,111. |
| 作者姓名： | 李明奎 王建丽 王明孝 齐建成 |
| 作者单位： | 1. 68029部队,甘肃兰州,730020 2. 甘肃省基础地理信息中心,甘肃兰州,730000 3. 解放军国际关系学院,江苏南京,210039 |
| 摘 要： | 传统混合像元分类算法按照同一标准同时提取各类地物,导致错分现象严重。本文利用端元光谱间的独立性关系和决策二叉树的分层特征,提出基于决策二叉树的混合像元分类方法。利用黄山地区Landsat7 ETM+影像数据进行实验,结果表明:该算法正确合理,分类精度比SGA-NCLS方法高,但分类效率比SGA-NCLS低。 |
| 关 键 词： | 决策二叉树 多光谱影像 混合像元分类 端元独立性 |
Mixed Pixel Classification of Multi-spectral Image Based on Binary Decision Tree |
|
| LI Mingkui,WANG Jianli,WANG Mingxiao,QI Jiancheng.Mixed Pixel Classification of Multi-spectral Image Based on Binary Decision Tree[J].Geospatial Information,2010,8(3):100-102,111. | |
| Authors: | LI Mingkui WANG Jianli WANG Mingxiao QI Jiancheng |
| Institution: | LI Mingkui1,WANG Jianli2,WANG Mingxiao1,QI Jiancheng3(1.68029 Troops,Lanzhou 730020,China,2.Gansu Provincial Center for Geomatics,Lanzhou 730000,3.Chinese People's Liberation International Studies University,Nanjing 210039,China) |
| Abstract: | Traditional mixed-pixel classification methods extract ground-object based on same criterion,which lead to serious misclassification.By multi-spectral independence and layer-feature of Binary Decision Tree(BDT),a new mixed-pixel classification method was proposed based on BDT.Landsat7 ETM+ image of Yellow Mountain area was choosed to testify this method,experimental results show that the classification accuracy of proposed method is better than SGA-NCLS,efficience is worse than SGA-NCLS. |
| Keywords: | binary decision tree(BDT) multi-spectral image mixed-pixel classification multi-spectral independence |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |