高级检索

基于光谱梯度角的高光谱影像流形学习降维法

向英杰, 杨桄, 张俭峰, 王琪

向英杰, 杨桄, 张俭峰, 王琪. 基于光谱梯度角的高光谱影像流形学习降维法[J]. 激光技术, 2017, 41(6): 921-926. DOI: 10.7510/jgjs.issn.1001-3806.2017.06.030
引用本文: 向英杰, 杨桄, 张俭峰, 王琪. 基于光谱梯度角的高光谱影像流形学习降维法[J]. 激光技术, 2017, 41(6): 921-926. DOI: 10.7510/jgjs.issn.1001-3806.2017.06.030
XIANG Yingjie, YANG Guang, ZHANG Jianfeng, WANG Qi. Dimensionality reduction for hyperspectral imagery manifold learning based on spectral gradient angles[J]. LASER TECHNOLOGY, 2017, 41(6): 921-926. DOI: 10.7510/jgjs.issn.1001-3806.2017.06.030
Citation: XIANG Yingjie, YANG Guang, ZHANG Jianfeng, WANG Qi. Dimensionality reduction for hyperspectral imagery manifold learning based on spectral gradient angles[J]. LASER TECHNOLOGY, 2017, 41(6): 921-926. DOI: 10.7510/jgjs.issn.1001-3806.2017.06.030

基于光谱梯度角的高光谱影像流形学习降维法

基金项目: 

吉林省教育厅"十二五"科研项目 2015448

吉林省科技发展计划资助项目 20140101213JC

详细信息
    作者简介:

    向英杰(1993-), 男, 硕士研究生, 主要研究方向为高光谱图像解译

    通讯作者:

    杨桄, E-mail:yg2599@126.com

  • 中图分类号: TP751.1

Dimensionality reduction for hyperspectral imagery manifold learning based on spectral gradient angles

  • 摘要: 为了挖掘高光谱数据的光谱局部特征,从高光谱遥感数据内在的非线性结构出发,提出了一种基于光谱梯度角的高光谱影像流形学习降维方法。采用局部化流形学习算法局部保持投影(LPP)对高光谱遥感数据进行非线性降维,对距离度量进行改进,将能够更好刻画高光谱影像光谱局部特征的光谱梯度角相似性度量应用于LPP方法,并用真实高光谱图像进行降维实验,取得了优于LPP方法和采用光谱角的LPP方法的结果。结果表明,在光谱规范化特征值方面,所提方法优于LPP方法和采用光谱角的LPP方法;在信息量的保持方面,具有更好的局部细节信息保持量。采用光谱梯度角的流形学习方法用于高光谱影像降维能取得较好的降维效果。
    Abstract: In order to extract the local characteristics of hyperspectral data, a dimensionality reduction method of hyperspectral imagery manifold learning based on spectral gradient angle was proposed from the nonlinear structure of hyperspectral imagery. Locality preserving projection (LPP) of localized manifold learning algorithm was performed to reduce the dimensionality of hyperspectral remote sensing data. In order to improve the distance metric, similarity measurement of spectral gradient angle, which can better characterize local features of hyperspectral images, was applied to LPP method. The real hyperspectral images were subjected to dimensionality reduction experiments.The results were better than the original LPP method and the LPP method using the spectral angle. The results show that the proposed method is superior to LPP method and LPP method using the spectral angle in the spectral normalized eigenvalues. Meanwhile, the proposed method can also obtain a good performance in information retainment and have better local information retention. Therefore, the manifold learning method with spectral gradient angle has a better performance in dimensionality reduction of hyperspectral images.
  • Figure  1.   The 20th band image of San Diego airport

    Figure  2.   The 20th band image of PAVIA university

    Figure  3.   Dimensionality reduction results of data 1 by three algorithms

    a—LPP algorithm b—SA-LPP algorithm c—SGA-LPP algorithm

    Figure  4.   Dimensionality reduction results of data 2 by three algorithms

    a—LPP algorithm b—SA-LPP algorithm c—SGA-LPP algorithm

    Figure  5.   Comparison of the normalized eigenspectra by LPP, SA-LPP and SGA-LPP

    a—data 1 b—data 2

    Figure  6.   RGB fusion results of data 1

    Figure  7.   RGB fusion results of data 2

    Table  1   Index evaluation of dimension reduction for data 1

    algorithms RRI MSE EI
    LPP 89.35% 1.7637×106 4.54
    SA-LPP 92.41% 1.4554×106 6.33
    SGA-LPP 94.54% 1.5469×106 6.98
    下载: 导出CSV

    Table  2   Index evaluation of dimension reduction for data 2

    algorithms RRI MSE EI
    LPP 88.42% 2.4735×106 5.24
    SA-LPP 90.87% 1.8373×106 7.85
    SGA-LPP 93.19% 1.9724×106 8.20
    下载: 导出CSV
  • [1]

    ZHANG B. The advances of hyperspectral image processing and information extraction[J]. Journal of Remote Sensing, 2016, 20(5):1061-1090(in Chinese). http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=b3ed5b32dedeff15b678360c472cbcbf

    [2]

    JOSÉ M, ANTONIO P, GUSTAVO C, et al. Hyperspectral remote sensing data analysis and future challenges[J]. IEEE Geoscience and Remote Sensing Magazine, 2013, 1(2):6-36. DOI: 10.1109/MGRS.2013.2244672

    [3]

    CHEN H D, PU H H, WANG B, et al. Image euclidean distance-based manifold dimensionality reduction algorithm for hyperspectral imagery[J]. Journal of Infrared and Millimeter Waves, 2013, 32(5):450-451(in Chinese). DOI: 10.3724/SP.J.1010.2013.00450

    [4]

    MULLER K R, MIKA S. An introduction in kernel-based learning algorithms[J]. IEEE Transactions on Neural Networks, 2001, 12(2):181-201. DOI: 10.1109/72.914517

    [5]

    BACHMANN C M, AINSWORTH T L, FUSINA R A. Exploiting manifold geometry in hyperspectral imagery[J]. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(3):441-454. DOI: 10.1109/TGRS.2004.842292

    [6]

    BACHMANN C M, AINSWORTH T L, FUSINA R A. Improved manifold coordinate representations of large scale hyperspectral scenes[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(10):2786-2803. DOI: 10.1109/TGRS.2006.881801

    [7]

    ZHANG A, XIE Y. Chaos theory-based data-mining technique for image endmember extraction: Laypunov index and correlation dimension (L and D) [J]. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(4):1935-1947. DOI: 10.1109/TGRS.2013.2256790

    [8]

    ROWEIS S T, SAUL L K. Nonlinear dimensionality reduction by locally linear embedding[J]. Science, 2000, 290(5500):2323-2326. DOI: 10.1126/science.290.5500.2323

    [9]

    TENENBAUM J, SILVA D D, LANGFORD J. A global geometric framework for nonlinear dimensionality reduction[J]. Science, 2000, 290(5500):2319-2323. DOI: 10.1126/science.290.5500.2319

    [10]

    MIKHAIL B, PARTH N. Laplacian eigenmaps for dimensionality reduction and data representation[J]. Neural Computation, 2003, 15(6):1373-1396. DOI: 10.1162/089976603321780317

    [11]

    HE X F, NIYOGI P. Locality preserving projections[J].Neural Information Processing System, 2004, 16:153-160. http://d.old.wanfangdata.com.cn/Periodical/rjxb201006008

    [12]

    DU P J, WANG X M, TAN K, et al. Dimension reduction and feature extraction of hyperspectral remote sensing imagery using manifold learning[J]. Geomatics and Information Science of Wuhan University, 2011, 36(2):148-152(in Chinese). http://en.cnki.com.cn/Article_en/CJFDTOTAL-WHCH201102005.htm

    [13]

    DING L, TANG P, LI H Y. Dimensionality reduction and classification for hyperspectral remote sensing data using ISOMAP[J]. Infrared and Laser Engineering, 2013, 42(10):2707-2711. http://en.cnki.com.cn/Article_en/CJFDTOTAL-HWYJ201310022.htm

    [14]

    WANG Y T, HUANG S Q, WANG H X, et al. Dimensionality reduction for hyperspectral image based on manifold learning[C]//International Conference on Image and Graphics. New York, USA: Springer International Publishing, 2015: 164-172.

    [15]

    WANG L L, LI Z Y, SUN J X, et al. New measure based manifold algorithm and application in anomaly detection of hyperspectral imagery[J]. Applied Mechanics and Materials, 2011, 80/81(3):797-803. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=10.4028/www.scientific.net/AMM.80-81.797

    [16]

    ZHANG X B, YUAN Y, WANG Q. Spectral discrimination method based on information divergence of gradient[J]. Acta Optica Sinica, 2011, 31(5):0530001(in Chinese). DOI: 10.3788/AOS

    [17]

    ZHANG Zh Y, ZHA H Y. Principal manifolds and nonlinear dimensionality reduction via tangent space alignment[J]. SIAM Journal of Scientific Computing, 2004, 26(1):313-338. http://d.old.wanfangdata.com.cn/Periodical/shdxxb-e200404005

图(7)  /  表(2)
计量
  • 文章访问数:  5
  • HTML全文浏览量:  1
  • PDF下载量:  2
  • 被引次数: 0
出版历程
  • 收稿日期:  2017-01-06
  • 修回日期:  2017-03-05
  • 发布日期:  2017-11-24

目录

    /

    返回文章
    返回