| 祝磊,胡奇峰,王棋林,杨君婷,严明.基于正交指数局部保留投影的高光谱图像特征提取[J].光电子激光,2019,30(9):968~977 |
| 基于正交指数局部保留投影的高光谱图像特征提取 |
| A hyperspectral image feature extraction method based on orthogonal exponential locality preserving projections |
| 投稿时间：2019-04-01 |
| DOI： |
| 中文关键词: 高光谱图像 特征提取 正交指数判别局部保留投影 |
| 英文关键词:hyperspectral image feature extraction orthogonal exponential discr iminant locality preserving projection |
| 基金项目:国家自然科学基金联合基金(U1609218)资助项目 (杭州电子科技大学 自动化学院,浙江 杭州 310018) |
|
| 摘要点击次数: 909 |
| 全文下载次数: 0 |
| 中文摘要: |
| 针对高光谱图像,在判别局部保留投影(Discri minant Locality Preserving Projection,DLPP)的基 础上,提出了一种名为正交指数判别局部保留投影(Orthogonal Exponential Discriminan t Locality Preserving Projection,OEDLPP)的特征提取方法。该算法不但保留了DLPP算法的有监督特性,还利用 了指数矩阵(the matrix exponential)来获取更有效的样本信息,避免了小样本问题。同时,OEDLPP 对投影矩阵进行 施密特正交化,解决了特征的冗余性问题。应用OEDLPP算法对高光谱图像进行特征提取后, 并采用支持 向量机(SVM)对降维后的数据进行分类。与主成分分析(PCA)、局部保留投影(LPP)、 判别局部保 留投影(DLPP)、指数判别局部保留投影(EDLPP)、正交判别局部保留投影(ODLPP)等对 比实验结 果表明,本文算法对样本有效信息的获取具有一定的优越性,分类精度提升了2%~3%左右。 |
| 英文摘要: |
| For hyperspectral images,we proposed a feature extraction method call ed orthogonal exponential discriminant locality preserving projection (OEDLPP) based on discri minant locality preserving projection (DLPP).The OEDLPP algorithms not only retains the supervi sed characteristic of the DLPP algorithms,but also utilizes the matrix exponential to obtain more effective sample information,which can avoid the Small Samples Size problem.Me anwhile, Schimidt orthogonalization is used to the projection matrix in OEDLPP,which sol ved the problems of feature redundancy.Last but not least,we used support vector machi ne (SVM) to classify the hyperspectral images after using OEDLPP algorithms to extract featu re.Compared with several existing algorithms,such as principal component analysis (PCA),loc ality preserving projection (LPP),discriminant locality preserving projection (DLPP),exponential discriminant locality preserving projection (EDLPP) and orthogonal discriminant locality pres erving projection (ODLPP),the proposed algorithms has a certain superiority for obtaini ng the effective information of the sample,and the classification accuracy is improved by about 2%~3%. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|