access icon free Graph-based spatial–spectral feature learning for hyperspectral image classification

Classifying hyperspectral data within high dimensionality is a challenging task. To cope with this issue, this study implements a semi-supervised multi-kernel class consistency regulariser graph-based spatial–spectral feature learning framework. For feature learning process, establishing the neighbouring relationship between the distinct samples from the high-dimensional space is the key to a favourable outcome for classification. The proposed method implements two kernels and a class consistency regulariser. The first kernel constructs simple edges where every single vertex represents one particular sample and the edge weight encodes the initial similarity between distinct samples. Later the obtained relation is fed into the second kernel to obtain the final features for classification where the semi-supervised learning is conducted to estimate the grouping relations among different samples according to their similarity, class, and spatial information. To validate the performance of proposed framework, the authors conduct several experiments on three publically available hyperspectral datasets. The proposed work equates favourably with state-of-the-art works with an overall classification accuracy of 98.54, 97.83, and 98.38% for Pavia University, Salinas-A, and Indian Pines datasets, respectively.

Inspec keywords: graph theory; feature extraction; geophysical image processing; image classification

Other keywords: hyperspectral image classification; Indian Pines datasets; classification feature; high-dimensional space; graph-based spatial-spectral feature learning; semisupervised multikernel class consistency regulariser; Pavia University; grouping relation; Salinas-A datasets; classification accuracy; initial similarity

Subjects: Algebra, set theory, and graph theory; Computer vision and image processing techniques; Geophysics computing; Combinatorial mathematics; Instrumentation and techniques for geophysical, hydrospheric and lower atmosphere research; Data and information; acquisition, processing, storage and dissemination in geophysics; Geophysical techniques and equipment; Combinatorial mathematics; Image recognition

References

    1. 1)
      • 1. Schott, J.R.: ‘Remote sensing: the image chain approach’ (Oxford University Press, 2007, 2nd edn.).
    2. 2)
      • 5. Ahmad, M., Lee, S.Y., Haq, I.U., et al: ‘Blind feature selection and extraction in 3-D image cube’, J. Flow Vis. Image Process., 2012, 19, (2), pp. 97111.
    3. 3)
      • 25. Gao, Y., Ji, R., Cui, P.: ‘Hyperspectral image classification through bilayer graph-based learning’, IEEE Trans. Image Process., 2014, 23, (7), pp. 27692778.
    4. 4)
      • 28. Liao, W., Pizurica, A., Bellens, R., et al: ‘Generalized graph based fusion of hyperspectral and LiDAR data using morphological features’, IEEE Geosci. Remote Sens. Lett., 2015, 12, (3), pp. 552556.
    5. 5)
      • 14. Cho, M., Lee, J., Lee, K.M.: ‘Reweighted random walks for graph matching’. Proc. of 11th European Conf. on Computer Vision, 2010, pp. 492505.
    6. 6)
      • 6. Medjahed, S.A., Saadi, T.A., Benyettou, A., et al: ‘Gray wolf optimizer for hyperspectral band selection’, J. Appl. Soft Comput., 2016, 40, pp. 178186.
    7. 7)
      • 2. Li, F., Xu, L., Siva, P., et al: ‘Hyperspectral image classification with limited labeled training samples using enhanced ensemble learning and conditional random fields’, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 2015, 8, (6), pp. 24272438.
    8. 8)
      • 23. Lawler, E.L.: ‘The quadratic assignment problem’, book part IV (NATO Advanced Study Institutes, Springer, 1974), vol. 19, pp. 351360.
    9. 9)
      • 10. Fauvel, M., Chanussot, J., Atli Bendiktsson, J., et al: ‘Kernel principal component analysis for the classification of hyperspectral remote sensing’, EURASIP J. Adv. Signal Process., 2009, 2009, pp. 114.
    10. 10)
      • 12. Plaza, A., Benediktsson, J.A., Boardman, J., et al: ‘Recent advances in techniques for hyperspectral image processing’, J. Remote Sens. Environ., 2009, 113, (1), pp. 110122.
    11. 11)
      • 29. Ahmad, M., Bashir, A.K., Khan, A.M.: ‘Metric similarity regularizer to enhance pixel similarity performance for hyperspectral unmixing’, J. Opt.-Int. J. Light Electron Opt., 2017, 140, pp. 8695.
    12. 12)
      • 38. Liao, W., Pizurica, A., Scheunders, P., et al: ‘Semi-supervised local discriminant analysis for feature extraction in hyperspectral images’, IEEE Trans. Geosci. Remote Sens., 2013, 51, (1), pp. 184198.
    13. 13)
      • 19. Leordeanu, M., Sukthankar, R., Hebert, M.: ‘Unsupervised learning for graph matching’, Int. J. Comput. Vis., 2012, 96, (1), pp. 2845.
    14. 14)
      • 9. Sellami, A., Farah, I.R.: ‘High-level hyperspectral image classification based on spectro-spatial dimensionality reduction’, J. Spat. Stat., 2016, 16, pp. 103117.
    15. 15)
      • 41. Luo, R., Liao, W., Huang, X., et al: ‘Feature extraction of hyperspectral images with semi-supervised graph learning’, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 2016, 9, (9), pp. 43894399.
    16. 16)
      • 24. Koopmans, T.C., Beckmann, M.: ‘Assignment problems and the location of economic activities’, Econometrica, 1957, 25, pp. 5376.
    17. 17)
      • 20. Gold, S., Rangarajan, A.: ‘A graduated assignment algorithm for graph matching’, IEEE Trans. Pattern Anal., 1996, 18, (4), pp. 377388.
    18. 18)
      • 31. Hou, B., Zhang, X., Ye, Q., et al: ‘A novel method for hyperspectral image classification based on Laplacian eigen map pixels distribution-flow’, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 2013, 6, (3), pp. 16021618.
    19. 19)
      • 35. Zhou, D., Bousquet, O., Lal, T.N., et al: ‘Learning with local and global consistency’, in Thrun, S., Saul, L.K., Sch, P.B. (Eds.): ‘Advances in neural information processing systems’ (MIT Press, Cambridge, USA, 2004), pp. 321328.
    20. 20)
      • 16. Eppstein, D.: ‘Subgraph isomorphism in planar graphs and related problems’. Proc. of 6th Annual ACM-SIAM Symp. on Discrete Algorithms, 1995, pp. 632640.
    21. 21)
      • 32. Nathan, N.D., Czaja, W., Messinger, D.d.W.: ‘Schrodinger eigen maps with non-diagonal potentials for spatial–spectral clustering of hyperspectral imagery’. Proc. of SPIE Algorithms and Technologies for Multispectral, Hyperspectral, and Ultra Spectral Imagery, 2014, vol. 9088.
    22. 22)
      • 22. Egozi, A., Keller, Y., Guterman, H.: ‘A probabilistic approach to spectral graph matching’, IEEE Trans. Pattern Anal., 2013, 35, (1), pp. 1827.
    23. 23)
      • 15. Leordeanu, M., Hebert, M., Sukthankar, R.: ‘An integer projected fixed point method for graph matching and map inference’. Proc. of Advanced Neural Information Processing and System, 2009, pp. 11141122.
    24. 24)
      • 7. Gong, M., Zhang, M., Yuan, Y.: ‘Unsupervised band selection based on evolutionary multi-objective optimization for hyperspectral images’, IEEE Trans. Geosci. Remote Sens., 2015, 54, (1), pp. 544557.
    25. 25)
      • 27. Liu, J., Wu, Z., Sun, L., et al: ‘Hyperspectral image classification using kernel sparse representation and semi-local spatial graph regularization’, IEEE Geosci. Remote Sens. Lett., 2013, 11, (8), pp. 13201324.
    26. 26)
      • 30. Gillis, D.B., Bowles, J.H.: ‘Hyperspectral image segmentation using spatial–spectral graphs’. Proc. of XVIII SPIE Conf. on Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery, 2012, 83901Q-11.
    27. 27)
      • 13. Zhou, F., Torre, F.D.: ‘Factorized graph matching’. Proc. of IEEE Conf. on Computer Vision and Pattern Recognition, 2012, pp. 127134.
    28. 28)
      • 4. Su, H., Yong, B., Du, Q.: ‘Hyperspectral band selection using improved firefly algorithm’, IEEE Geosci. Remote Sens. Lett., 2015, 13, (1), pp. 6872.
    29. 29)
      • 37. Sugiyama, M., Lde, T., Nakajima, S., et al: ‘Semi-supervised local Fisher discriminant analysis for dimensionality reduction’, Mach. Learn., 2009, 78, pp. 3561.
    30. 30)
      • 39. Luo, R., Liao, W., Philips, W., et al: ‘An improved semi-supervised local discriminant analysis for feature extraction of hyperspectral image’. Proc. of Joint Urban Remote Sensing and Event (JURSE), 2015, pp. 14.
    31. 31)
      • 17. Luks, E.M.: ‘Isomorphism of graphs of bounded valence can be tested in polynomial time’, J. Comput. Syst. Sci., 1982, 25, pp. 4265.
    32. 32)
      • 33. Czaja, W., Ehler, M.: ‘Schrodinger eigen maps for the analysis of biomedical data’, IEEE Trans. Pattern Anal. Mach. Intell., 2013, 35, (5), pp. 12741280.
    33. 33)
      • 40. Zhang, S., Yu, G.: ‘Semi-supervised locality preserving projections with compactness enhancement’. Proc. of Int. Conf. on Education of Information Technology (ICEIT), 2010, pp. 460464.
    34. 34)
      • 3. Jia, S., Tang, G., Zhu, J., et al: ‘A novel ranking-based clustering approach for hyperspectral band selection’, IEEE Trans. Geosci. Remote Sens., 2015, 54, (1), pp. 88102.
    35. 35)
      • 26. Bencherif, M.A., Bazi, Y., Guessoum, A., et al: ‘Fusion of extreme learning machine and graph-based optimization methods for active classification of remote sensing images’, IEEE Geosci. Remote Sens. Lett., 2014, 12, (3), pp. 527531.
    36. 36)
      • 8. Ahmad, M., Haq, I.U., Mushtaq, Q., et al: ‘A new statistical approach for band clustering and band selection using K-means clustering’, Int. J. Eng. Technol., 2011, 3, (6), pp. 606614.
    37. 37)
      • 11. Ahmad, M., Haq, I.U., Mushtaq, Q.: ‘AIK method for band clustering using statistics of correlation and dispersion matrix’. Proc. of IPCSIT, 2011, vol. 16, pp. 114118.
    38. 38)
      • 36. Camps-Valls, G., Bandos, T.V., Zhou, D.: ‘Semi-supervised graph based hyperspectral image classification’, IEEE Trans. Geosci. Remote Sens., 2007, 45, (10), pp. 30443054.
    39. 39)
      • 21. Caetano, T.S., Cheng, L., Le, Q.V., et al: ‘Learning graph matching’, IEEE Trans. Pattern Anal., 2009, 31, (8), pp. 10481058.
    40. 40)
      • 18. Aho, A.V., Ganapathi, M., Tiang, S.W.: ‘Code generation using tree matching and dynamic programming’, ACM Trans. Program Lang. Syst., 1989, 11, pp. 491516.
    41. 41)
      • 34. Publicly available data sets. Available at http://www.ehu.eus/ccwintco/index.php?title=Hyperspectral-Remote-Sensing-Scenes, 2016.
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-ipr.2017.0168
Loading

Related content

content/journals/10.1049/iet-ipr.2017.0168
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading