Please wait a minute...
浙江大学学报(工学版)
JOURNAL OF ZHEJIANG UNIVERSITY (ENGINEERING SCIENCE)  2018, Vol. 52 Issue (4): 687-693    DOI: 10.3785/j.issn.1008-973X.2018.04.011
Automatic Technology     
Improved kernel symmetric sparse representation based band selection for hyperspectral imagery
SUN Wei-wei1,2, MA Jun1, YANG Gang1, LI Wei-yue3,4
1. Laboratory of Geography Information Techniques, Ningbo University, Ningbo 315211, China;
2. State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China;
3. Institute of Urban Studies, Shanghai Normal University, Shanghai 200234, China;
4. Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences, Lanzhou 730000, China
Download:   PDF(1854KB) HTML
Export: BibTeX | EndNote (RIS)      

Abstract  

An improved kernel symmetric sparse representation (IKSSR) method was proposed to make dimensionality reduction and select a proper band subset from hyperspectral imagery (HSI). The IKSSR adopted the kernel function and binary constraints in the sparse coefficient matrix in order to improve the regular symmetric sparse representation (SSR). The convex hull in high dimensional parametric space was constructed to contain the corresponding points of all band vectors in the HSI data, and a proper band subset which is transformed into finding the archetypes of its minimal convex hull was selected. The IKSSR implemented the vector quantization scheme to initialize the band subset and optimized the convex sub-problems via the block coordinate descend method to obtain the desired band subset. Two open hyperspectral datasets were implemented to testify the performance of IKSSR and the results on classification were compared against four state-of-the-art methods. Experimental results show that the proposed IKSSR achieves better overall classification accuracy (OCA) results than the regular SSR and other three methods.

Received: 12 November 2016     
CLC:  P237  
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
Cite this article:

SUN Wei-wei, MA Jun, YANG Gang, LI Wei-yue. Improved kernel symmetric sparse representation based band selection for hyperspectral imagery. JOURNAL OF ZHEJIANG UNIVERSITY (ENGINEERING SCIENCE), 2018, 52(4): 687-693.

URL:

http://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2018.04.011     OR     http://www.zjujournals.com/eng/Y2018/V52/I4/687


改进核空间对称稀疏表达用于高光谱波段选择

提出改进核空间对称稀疏表达（IKSSR）降维方法来解决高光谱影像（HSI）的波段选择问题.该方法利用核函数方法和稀疏系数的二值约束条件改进对称稀疏表达模型，在映射得到的核空间构建包含所有波段向量对应的数据点的凸包，通过寻找最小凸包的原型点实现波段子集的优化选择.改进核空间对称稀疏表达方法采用矢量量化策略初始化波段子集，利用块坐标下降方法将非凸问题转换为凸目标函数优化问题，实现目标波段子集选取.基于两个公开高光谱数据集，对该方法和4种主流的波段选择方法进行实验比较研究.实验结果表明，利用改进核空间的对称稀疏表达方法得到的总体分类精度优于对称稀疏表达模型和其他3种方法.

[1] TONG Q, XUE Y, ZHANG L. Progress in hyperspectral remote sensing science and technology in China over the past three decades[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2014, 7(1):70-91.
[2] BIOUCAS-DIAS J M, PLAZA A, CAMPS-VALLS G, et al. Hyperspectral remote sensing data analysis and future challenges[J]. IEEE Geoscience and remote sensing magazine, 2013, 1(2):6-36.
[3] SUN W, JIANG M, LI W, et al. A symmetric sparse representation based band selection method for hyperspectral imagery classification[J]. Remote Sensing, 2016, 8(3):238.
[4] 孙伟伟, 蒋曼, 李巍岳. 利用稀疏自表达实现高光谱影像波段选择[J]. 武汉大学学报信息科学版, 2017, 42(4):441-448. SUN Wei-wei, JIANG Man, LI Wei-yue. Band selection using sparse self-representation for hyperspectral imagery[J]. Geomatics and Information Science of Wuhan University, 2017,42(4):441-448.
[5] LI J M, QIAN Y T. Clustering-based hyperspectral band selection using sparse nonnegative matrix factorization[J]. Journal of Zhejiang University:Science C, 2011, 12(7):542-549.
[6] CHEPUSHTANOVA S, GITTINS C, KIRBY M. Band selection in hyperspectral imagery using sparse support vector machines[C]//Proceedings of SPIE Defense+ Security. Baltimore:SPIE, 2014:90881F-90881F-15.
[7] DONOHO D L. Compressed sensing[J]. IEEE Transactions on Information Theory, 2006, 52(4):1289-1306.
[8] 厉小润, 伍小明, 赵辽英. 非监督的高光谱混合像元非线性分解方法[J]. 浙江大学学报:工学版, 2011, 45(4):607-613. LI Xiao-run, WU Xiao-ming, ZHAO Liao-ying. Unsupervised nonlinear decomposing method of hyperspectral imagery[J]. Journal of Zhejiang University:Engineering Science, 2011, 45(4):607-613.
[9] WEI Q, BIOUCAS-DIAS J, DOBIGEON N, et al. Hyperspectral and multispectral image fusion based on a sparse representation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2015, 53(7):3658-3668.
[10] LUO F, HUANG H, MA Z, et al. Semisupervised sparse manifold discriminative analysis for feature extraction of hyperspectral images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(10):6197-6211.
[11] XU Y, WU Z, LI J, et al. Anomaly detection in hyperspectral images based on low-rank and sparse representation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(4):1990-2000.
[12] ZHANG E, ZHANG X, JIAO L, et al. Spectral-spatial hyperspectral image ensemble classification via joint sparse representation[J]. Pattern Recognition, 2016, 59(1):42-54.
[13] 厉小润, 朱洁尔, 王晶, 等. 组合核支持向量机高光谱图像分类[J]. 浙江大学学报:工学版, 2013, 47(8):1403-1410. LI Xiao-run, ZHU Jie-er, WANG Jing, et al. Hyperspectral image classification based on composite kernels support vector machine[J]. Journal of Zhejiang University:Engineering Science, 2013, 47(8):1403-1410.
[14] DU Q, BIOUCAS-DIAS J M, PLAZA A. Hyperspectral band selection using a collaborative sparse model[C]//Proceedings of 2012 IEEE International Geoscience and Remote Sensing Symposium. Munich:IEEE, 2012:3054-3057.
[15] SUN W, ZHANG L, DU B, et al. Band selection using improved sparse subspace clustering for hyperspectral imagery classification[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015, 8(6):2784-2797.
[16] CUTLER A, BREIMAN L. Archetypal analysis[J]. Technometrics, 1994, 36(4):338-347.
[17] CHEN Y, MAIRAL J, HARCHAOUI Z. Fast and robust archetypal analysis for representation learning[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus:IEEE, 2014:1478-1485.
[18] MORUP M, HANSEN L K. Archetypal analysis for machine learning and data mining[J]. Neurocomputing, 2012, 80(1):54-63.
[19] KWAK N. Nonlinear projection trick in Kernel methods:an alternative to the Kernel trick[J]. IEEE Transactions on Neural Networks and Learning Systems, 2013, 24(12):2113-2119.
[20] EQUITZ W H. A new vector quantization clustering algorithm[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989, 37(10):1568-1575.
[21] NOCEDAL J, WRIGHT S. Numerical optimization[M]. 2nd ed. NewYork:Springer, 2006:110.
[22] CHANG C I, DU Q, SUN T L, et al. A joint band prioritization and band-decorrelation approach to band selection for hyperspectral image classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 1999, 37(6):2631-2641.
[23] LI S, QI H. Sparse representation based band selection for hyperspectral images[C]//Proceedings of 201118th IEEE International Conference on Image Processing. Brussels:IEEE, 2011:2693-2696.
[1] GUO Zhong-hao, REN Yu-peng, QIN Yi, WANG Xin, GU Juan, MA Jing-yu, ZOU Le-jun, SHEN Xiao-hua. Cloud detection method based on scatter plot for Landsat 8 image with adjustable threshold[J]. JOURNAL OF ZHEJIANG UNIVERSITY (ENGINEERING SCIENCE), 2018, 52(7): 1423-1430.