Dimensionality Reduction of Hyperspectral Data via Spectral Feature Extraction

被引:76
作者
Mojaradi, Barat [1 ,2 ]
Abrishami-Moghaddam, Hamid [3 ,4 ]
Zoej, Mohammad Javad Valadan [1 ]
Duin, Robert P. W. [5 ]
机构
[1] KN Toosi Univ Technol, Fac Geodesy, Tehran 16785163, Iran
[2] Shahid Rajaee Univ, Tehran 16785163, Iran
[3] KN Toosi Univ Technol, Fac Elect Engn, Tehran 163151355, Iran
[4] Univ Picardie, Dept Biophys, F-80000 Amiens, France
[5] Delft Univ Technol, Elect Engn Math & Comp Sci Dept, NL-2600 GA Delft, Netherlands
来源
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING | 2009年 / 47卷 / 07期
关键词
Feature extraction (FE); hyperspectral images; remote-sensing image classification; spectral feature; FEATURE-SELECTION; BAND SELECTION; CLASSIFICATION; ALGORITHMS;
D O I
10.1109/TGRS.2008.2010346
中图分类号
P3 [地球物理学]; P59 [地球化学];
学科分类号
0708 ; 070902 ;
摘要
This paper proposes an innovative spectral feature extraction (SFE) method called prototype space (PS) feature extraction (PSFE) based only on class spectra. The main novelties of the proposed SFE lie in the following: representing channels in a new space called PS, where they are characterized in terms of reflection properties of classes; and proposing uncertainty, angle, and distance measures to distinguish highly correlated and informative channels. Having clustered the channels by Fuzzy C-Means (FCM) in PS, highly correlated and isolated channels are separated by an uncertainty measure. Consequently, PSFE is built by a linear combination of spectra weighted by class membership values of channels that fall in each cluster. Furthermore, we will enrich PSFE with informative channels which are outlier channels identified through their angle and distance with respect to diagonal and cluster centers in PS. In contrast to the previous SFE methods, PSFE substitutes the search strategies with FCM clustering to find relevant channels. Moreover, instead of optimizing separability criteria, the accuracy of classification over a subset of training data is used to decide which disjoint optical region yields maximum accuracy. According to how class spectra are obtained, PSFE incorporates four approaches: knowledge based, supervised, semisupervised, and unsupervised. The latter three PSFE approaches are assessed in two main cases including with and without informative channels and compared with the conventional feature extraction methods. Experimental results demonstrated higher overall accuracy of PSFE compared to its conventional counterparts with limited sample sizes.
引用
收藏
页码:2091 / 2105
页数:15
相关论文
共 34 条
[1]   FCM - THE FUZZY C-MEANS CLUSTERING-ALGORITHM [J].
BEZDEK, JC ;
EHRLICH, R ;
FULL, W .
COMPUTERS & GEOSCIENCES, 1984, 10 (2-3) :191-203
[2]   Nonparametric discriminant analysis and nearest neighbor classification [J].
Bressan, M ;
Vitrià, J .
PATTERN RECOGNITION LETTERS, 2003, 24 (15) :2743-2749
[3]   Constrained band selection for hyperspectral imagery [J].
Chang, Chein-I ;
Wang, Su .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2006, 44 (06) :1575-1585
[4]  
Duin R., 2004, PRTOOLS MATLAB TOOLB
[5]  
Fukunaga K., 1990, Introduction to Statistical Pattern Recognition, DOI [10.5555/92131, DOI 10.5555/92131]
[6]  
Greenberger D.B., 1988, Journal of small business management, V26, P1
[7]  
Guyon I., 2006, Stud Fuzziness Soft Comput
[8]   Investigation of the random forest framework for classification of hyperspectral data [J].
Ham, J ;
Chen, YC ;
Crawford, MM ;
Ghosh, J .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2005, 43 (03) :492-501
[9]   Band selection based on feature weighting for classification of hyperspectral data [J].
Huang, R ;
He, MY .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2005, 2 (02) :156-159
[10]   Independent component analysis:: algorithms and applications [J].
Hyvärinen, A ;
Oja, E .
NEURAL NETWORKS, 2000, 13 (4-5) :411-430