| 基于非线性判别分析的高光谱影像特征提取 |
| |
| 引用本文： | 杨国鹏,余旭初.基于非线性判别分析的高光谱影像特征提取[J].测绘科学,2007,32(5):91-93. |
| |
| 作者姓名： | 杨国鹏 余旭初 |
| |
| 作者单位： | 解放军信息工程大学测绘学院,郑州,450052;解放军信息工程大学测绘学院,郑州,450052 |
| |
| 摘 要： | 高光谱遥感影像丰富的光谱信息有利于深入挖掘目标的理化特性,精细识别不同目标间的细微差异。为了提高影像分类识别的精度与速度需要对光谱信息进行特征提取。基于核函数的判别分析能够在数据中提取非线性特征,本文将其应用到高光谱影像分类的特征提取中,并进行了最小距离分类实验取得理想结果。
|
| 关 键 词： | 高光谱影像 特征提取 判别分析 核函数 分类 |
| 文章编号： | 1009-2307（2007）05-0091-03 |
| 修稿时间： | 2006-05-30 |
Hyperspectral image feature extraction based on nonlinear discriminate analysis |
| |
| YANG Guo-peng,YU Xu-chu.Hyperspectral image feature extraction based on nonlinear discriminate analysis[J].Science of Surveying and Mapping,2007,32(5):91-93. |
| |
| Authors: | YANG Guo-peng YU Xu-chu |
| |
| Abstract: | There is abundant spectral information in hyperspectral remote sensing image,it can help people to mine feature of target,to recognize the imperceptible difference between various targets.In order to improve the precision and speed of classification in hyperspectral image,it is needed to extract feature of spectral information.Kernel-based discriminate analysis can be used to extract the features in nonlinear data,so it is used in the field of hyperspectral processing.The test indicates that it's a good method. |
| |
| Keywords: | hyperspectral image feature extraction discriminate analysis kernel function classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
