| 面向分类的高光谱影像局部线性嵌入算法研究 |
| |
| 引用本文： | 杨明,余旭初,吴翰书,王超,林斌.面向分类的高光谱影像局部线性嵌入算法研究[J].测绘科学,2012(4):29-31. |
| |
| 作者姓名： | 杨明 余旭初 吴翰书 王超 林斌 |
| |
| 作者单位： | 信息工程大学测绘学院;96215部队;深圳大学 |
| |
| 摘 要： | 局部线性嵌入算法(LLE)能很好保存数据点的局部性质,因此有很好的数据可视化效果,但它不是一种很好的面向分类的特征提取方法。因为它存在样本外点学习能力差和忽略了样本类别信息的缺点。对此,本文提出一种分类型局部线性嵌入算法。所提方法通过计算重构误差来判定样本类别,并引进平移向量和缩放因子对距离修正,显著提高类别可分性。在对高光谱影像进行分类的试验中验证了该方法的有效性。
|
| 关 键 词： | 局部线性嵌入 分类型局部线性嵌入 重构误差 缩放因子 |
Classification-oriented locally linear embedding algorithm of hyperspectral imagery |
| |
| YANG Ming,YU Xu-chu,WU Han-shu,WANG Chao,LIN Bin.Classification-oriented locally linear embedding algorithm of hyperspectral imagery[J].Science of Surveying and Mapping,2012(4):29-31. |
| |
| Authors: | YANG Ming YU Xu-chu WU Han-shu WANG Chao LIN Bin |
| |
| Institution: | ①(①Institute of Surveying and Mapping,Information Engineering University,Zhengzhou 450052,China;②Troops 96215,Guangxi Liuzhou 545616,China;③ShenZhen University,Guangdong Shenzhen 518060,China) |
| |
| Abstract: | Locally Linear Embedding(LLE) can preserve solely local properties of the data,so it facilitates the visualization of high-dimensional data,but it isn′t a good feature extraction method on classification.Because it is not good at sample outer point′s learning and the class labels are not used.In the paper,a classification based on locally linear embedding method was proposed to solve the problem.To improve classification result,the reconstruction error was used to classify the test sample and extension factor to verify object distance.To verify the effectiveness of the proposed method,the experiment was conducted on hyperspectral imagery classification and the result showed the effectiveness of the proposed method. |
| |
| Keywords: | locally linear embedding(LLE) classification-oriented locally linear embedding reconstruction error extension factor |
| 本文献已被 CNKI 等数据库收录！ |
|
