论文:2012,Vol:30,Issue(1):138-144
引用本文:
温金环, 田铮, 林伟, 周敏, 延伟东. 基于近邻保留PNMF特征提取的高光谱图像分类[J]. 西北工业大学
Wen Jinhuan, Tian Zheng, Lin Wei, Zhou Min, Yan Weidong. A New and Effective Method of NPPNMF (Neighborhood Preserving Projective Nonnegative Matrix Factorization) Feature Extraction for Hyperspectral Image Classification[J]. Northwestern polytechnical university

基于近邻保留PNMF特征提取的高光谱图像分类
温金环, 田铮, 林伟, 周敏, 延伟东
西北工业大学 应用数学系,陕西 西安 710072
摘要:
通过对投影非负矩阵分解(PNMF)增加近邻保留假设,提出了一种新的高光谱图像线性特征提取方法———近邻保留投影非负矩阵分解(NPPNMF)。NPPNMF保留了高光谱数据在低维特征空间中的局部几何结构,克服了PNMF基于Euclidean的缺点。根据在构造k近邻图时是否使用训练样本的类标签信息决定了NPPNMF既可以是无监督的特征提取方法,也可以是有监督的特征提取方法,从而提高了PNMF算法的鉴别力。理论证明和高光谱图像数据的分类结果表明了该方法的有效性及应用潜力。
关键词:    高光谱图像分类    特征提取    降维    投影非负矩阵分解    近邻保留   
A New and Effective Method of NPPNMF (Neighborhood Preserving Projective Nonnegative Matrix Factorization) Feature Extraction for Hyperspectral Image Classification
Wen Jinhuan, Tian Zheng, Lin Wei, Zhou Min, Yan Weidong
Department of Applied Mathematics,Northwestern Polytechnical University,Xi'an 710072,China
Abstract:
Aim.The introduction of the full paper reviews relevant matters and then proposes the NPPNMF featureextraction method mentioned in the title,which we believe is new and effective and which is explained in sections 1through 3.Our new method incorporates the neighborhood preserving assumption.Section 1 briefs PNMF method.Section 2 explains our new NPPNMF method; it is divided into subsections 2.1 and 2.2.Section 3 analyzes theconvergence of NPPNMF method; it gives the proof of four theorems.Section4 deals with experimental results andtheir analysis.Subsection 4.1 briefs AVIRIS hyperspectral data set.Subsection 4.3 presents experimental results onsuch data set in Figs.2 and 3 and Table 1 and analyze these results.The theoretical analysis in section 3 and the a-nalysis of experimental results in section 4 demonstrate preliminarily that the proposed new method is effective andpromising in hyperspectral image classification.
Key words:    analysis    classification (of information)    data mining    efficiency    experiments    feature extraction    im-age analysis    image processing    matrix algebra    pattern recognition    dimensionality reduction    hyper-spectral image    NPPNMF (Neighborhood Preserving Projective Nonnegative Matrix Factorization)   
收稿日期: 2011-04-02     修回日期:
DOI:
基金项目: 国家自然科学基金(10926197、60972150);陕西省自然科学基金(2010JQ1015);西北工业大学基础研究基金(JC201053)资助
通讯作者:     Email:
作者简介: 温金环(1974-),女,西北工业大学讲师、博士研究生,主要从事流形学习、非负矩阵分解及高光谱图像处理研究。
相关功能
PDF(861KB) Free
打印本文
把本文推荐给朋友
作者相关文章
温金环  在本刊中的所有文章
田铮  在本刊中的所有文章
林伟  在本刊中的所有文章
周敏  在本刊中的所有文章
延伟东  在本刊中的所有文章

参考文献:
[1] Jimenez L O,Rivera-Median J L,Rodirguez-Diaz E. Integration of Spatial and Spectral Information by Means of UnsupervisedExtraction and Classification for Homogenous Objects Applied to Multispectral and Hyperspectral Data. IEEE Trans on Geosci-ence and Remote Sensing, 2005, 43:844~851
[2] Rodarmel C,Shan J. Principal Component Analysis for Hyperspectral Image Classification. Surveying and Land Information Sys-tems, 2002, 62:115~122
[3] Jia X P,Richards J A. Segemted Principal Comonpents Transformation for Efficient Hyperspectral Remote Sensing Image Displayand Classification. IEEE Trans on Geoscience and Remote Sensing, 1999, 37(1): 538 -542
[4] Ren J M, Yu X C, Hao B X. Classification of Landsat TM Image Based on Non-Negative Matrix Factorization. IEEE Internation-al Geoscience and Remote Sensing Symposium 2007, 405~408
[5] Lee D,Seung H. Learning the Parts of Objects by Nonnegative Matrix Factorization. Nature, 1999, 401: 788~791
[6] Lee D,Seung H S. Algorithms for Non-Negatvie Matrix Factorization. Advances in Neural Information Processing Systems,2001, 13:1~8
[7] Li S, Hou X, Zhang H, Cheng Q. Learning Spatially Localized, Parts Based Representation. IEEE Computer Society Conferenceon Computer Vision and Pattern Recognition, 2001, 1:207~212
[8] Lee H Y,Yoo J,Choi S J. Semi-Supervised Nonnegative Matrix Factorization. IEEE Trans on Signal Processing Letters, 2010,1(9): 4~7
[9] Chris D,Li T,Orthogonal Nonnegative Matrix Tri-Factorization for Clustering. Proceedings of the 12 th ACMSIGKDD Internation-al Conference on Knowledge Discovery and Data Mining, 2006,Philadelphia,USA:ACM, 2006: 126~135
[10] Liu X B,Yan S C,Jin H. Projective Nonnegative Graph Embedding. IEEE Trans on Image Processing,2010,19(5):1126~1137
[11] 李 乐,章毓晋,基于线性投影结构的非负矩阵分解. 自动化学报, 2010, 36(1):23~39Li Le,Zhang Yujin. Linear Projection-Based Non-Negative Matrix Factorization. Acta Automatica Sinica, 2010, 36(1):23~39(in Chinese)
[12] Roweis S T,Saul L K. Nonlinear Dimensionality Reduction By Locally Linear Embedding. Science, 2000, 290: 2323~2326
[13] He X F,Cai D. Neighborhood Preserving Embedding. Proc of 10th IEEE Internat Conf on Computer Vision, 2005: 1208~1213
[14] Cai D,He X F,Han J W. Graph Regularized Non-Negative Matrix Factorization for Data Representation. UIUC Computer Sci-ence Research and Tech Reports, 2008
[15] Yang Z R,Erkki O. Linear and Nonlinear Projection Non-Negatvie Matrix Factorization. IEEE Trans on Neural Networks, 2010,21:734~749
相关文献:
1.温金环, 田铮, 赵永强, 延伟东.基于局部保持线性判别嵌入特征提取的光谱图像分类[J]. 西北工业大学, 2013,31(2): 323-328