| 简化模糊H-ARTMAP神经网络在高光谱遥感分类中的应用 |
| |
| 引用本文： | 张伟,杜培军,尹作霞.简化模糊H-ARTMAP神经网络在高光谱遥感分类中的应用[J].测绘科学,2008,33(5). |
| |
| 作者姓名： | 张伟 杜培军 尹作霞 |
| |
| 作者单位： | 中国矿业大学,地理信息与遥感科学系,江苏徐州,221008;中国矿业大学,地理信息与遥感科学系,江苏徐州,221008;中国矿业大学,地理信息与遥感科学系,江苏徐州,221008 |
| |
| 基金项目： | 国家自然科学基金，国家重点实验室基金 |
| |
| 摘 要： | 人工神经网络作为一种不需估计类别分布参数的遥感影像分类方法,能够克服分类中的不确定性,提高分类精度。模糊ARTMAP人工神经网络具有稳定、泛化性能好、支持增量式学习等特点,通过对简化模糊AR-TMAP神经网络和H-ARTMAP神经网络的分析和集成,构造了一种用于高光谱遥感影像分类的简化模糊H-ART-MAP网络。实验证明该方法在分类效率、运算时间和分类精度等方面都优于最大似然分类、BP神经网络、最小距离分类、光谱角制图模型等分类方法。
|
| 关 键 词： | 简化模糊ARTMAP神经网络 H-ARTMAP神经网络 超球体 高光谱遥感 分类 |
Application of simplified fuzzy H-ARTMAP network to hyperspectral remote sensing classification |
| |
| ZHANG Wei,DU Pei-jun,YIN Zuo-xia.Application of simplified fuzzy H-ARTMAP network to hyperspectral remote sensing classification[J].Science of Surveying and Mapping,2008,33(5). |
| |
| Authors: | ZHANG Wei DU Pei-jun YIN Zuo-xia |
| |
| Abstract: | Artificial Neural Networks have been widely used in remote sensing classification and gained higher accuracy than traditional statistical classifiers since they don't need to estimate probability distribution and parameters.Simplified Fuzzy H-ARTMAP Network,that combines the advantages of Fuzzy ARTMAP and H-ARTMAP Network,is used to construct a classifier for hyperspectral remote sensing imagery.Experiments demonstrate that Simplified Fuzzy H-ARTMAP Network outperforms those traditional methods,e.g.MLC,BP Network,Minimum Distance,Mahalanobis Distance,Spectral Angle Mapping,Binary Encoding classification,and it has higher efficiency than BPNN due to less processing time and computation burden. |
| |
| Keywords: | simplified fuzzy ARTMAP neural network H-ARTMAP neural network hypersphere hyperspectral remote sensing classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
