Hyperspectral Dimensionality Reduction by Tensor Sparse and Low-Rank Graph-Based Discriminant Analysis

被引:47
作者
Pan, Lei [1 ]
Li, Heng-Chao [1 ]
Deng, Yang-Jun [1 ]
Zhang, Fan [2 ]
Chen, Xiang-Dong [1 ]
Du, Qian [3 ]
机构
[1] Southwest Jiaotong Univ, Sch Informat Sci & Technol, Chengdu 610031, Peoples R China
[2] Beijing Univ Chem Technol, Coll Informat Sci & Technol, Beijing 100029, Peoples R China
[3] Mississippi State Univ, Dept Elect & Comp Engn, Starkville, MS 39762 USA
基金
中国国家自然科学基金;
关键词
hyperspectral image; sparse and low-rank graph; tensor; dimensionality reduction; REPRESENTATION; CLASSIFICATION; FRAMEWORK;
D O I
10.3390/rs9050452
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Recently, sparse and low-rank graph-based discriminant analysis (SLGDA) has yielded satisfactory results in hyperspectral image (HSI) dimensionality reduction (DR), for which sparsity and low-rankness are simultaneously imposed to capture both local and global structure of hyperspectral data. However, SLGDA fails to exploit the spatial information. To address this problem, a tensor sparse and low-rank graph-based discriminant analysis (TSLGDA) is proposed in this paper. By regarding the hyperspectral data cube as a third-order tensor, small local patches centered at the training samples are extracted for the TSLGDA framework to maintain the structural information, resulting in a more discriminative graph. Subsequently, dimensionality reduction is performed on the tensorial training and testing samples to reduce data redundancy. Experimental results of three real-world hyperspectral datasets demonstrate that the proposed TSLGDA algorithm greatly improves the classification performance in the low-dimensional space when compared to state-of-the-art DR methods.
引用
收藏
页数:20
相关论文
共 37 条
[1]   Dimensionality Reduction Based on Group-Based Tensor Model for Hyperspectral Image Classification [J].
An, Jinliang ;
Zhang, Xiangrong ;
Jiao, L. C. .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2016, 13 (10) :1497-1501
[2]  
[Anonymous], 2002, Principal components analysis
[3]   Classification of Hyperspectral Images With Regularized Linear Discriminant Analysis [J].
Bandos, Tatyana V. ;
Bruzzone, Lorenzo ;
Camps-Valls, Gustavo .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2009, 47 (03) :862-873
[4]   Classification of hyperspectral data from urban areas based on extended morphological profiles [J].
Benediktsson, JA ;
Palmason, JA ;
Sveinsson, JR .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2005, 43 (03) :480-491
[5]   Distributed optimization and statistical learning via the alternating direction method of multipliers [J].
Boyd S. ;
Parikh N. ;
Chu E. ;
Peleato B. ;
Eckstein J. .
Foundations and Trends in Machine Learning, 2010, 3 (01) :1-122
[6]   A SINGULAR VALUE THRESHOLDING ALGORITHM FOR MATRIX COMPLETION [J].
Cai, Jian-Feng ;
Candes, Emmanuel J. ;
Shen, Zuowei .
SIAM JOURNAL ON OPTIMIZATION, 2010, 20 (04) :1956-1982
[7]   Progressive dimensionality reduction by transform for hyperspectral imagery [J].
Chang, Chein-I ;
Safavi, Haleh .
PATTERN RECOGNITION, 2011, 44 (10-11) :2760-2773
[8]   A unified semi-supervised dimensionality reduction framework for manifold learning [J].
Chatpatanasiri, Ratthachat ;
Kijsirikul, Boonserm .
NEUROCOMPUTING, 2010, 73 (10-12) :1631-1640
[9]  
Cheng B., 2015, IEEE T IMAGE PROCESS, V23, P2241
[10]   Morphological Attribute Profiles for the Analysis of Very High Resolution Images [J].
Dalla Mura, Mauro ;
Benediktsson, Jon Atli ;
Waske, Bjoern ;
Bruzzone, Lorenzo .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2010, 48 (10) :3747-3762