ISSN 2070-7401 (Print), ISSN 2411-0280 (Online)
Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa


Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2019, Vol. 16, No. 1, pp. 46-57

Development of the method of spectral-spatial classification of hyperspectral images based on local multifractal analysis and the support vector machine

Dm.V. Uchaev 1 
1 Moscow State University of Geodesy and Cartography, Moscow, Russia
Accepted: 08.11.2018
DOI: 10.21046/2070-7401-2019-16-1-46-57
Today, improving the quality of classification of hyperspectral images is one of the main problems of the theory and practice of hyperspectral image processing. One way to solve this problem is to involve spatial (contextual) information in the process of classification. In the paper, a new method of spectral-spatial classification of hyperspectral images is proposed. In the method, the multifractal characteristics and values of the Choquet capacity, calculated using local multifractal analysis, are used as spatial features. The paper identifies three groups of multifractal characteristics that can be used as classification features: global multifractal characteristics calculated for relatively large pixel neighborhoods; characteristics derived from global multifractal characteristics and local multifractal characteristics. The combination of spatial features with spectral information in the proposed method is performed using the support vector machine classifier. Experiments are carried out with two hyperspectral images of Pavia University and Salinas, having different spatial resolution. Experimental results demonstrate that the proposed method outperforms similar methods in terms of the overall accuracy and the kappa statistic.
Keywords: hyperspectral image, support vector machine, Holder exponent, multifractal
Full text


  1. Malinnikov V. A., Uchaev D. V., Primenenie metodiki mul’tifraktal’noi segmentatsii izobrazhenii dlya vydeleniya konturov na aerokosmicheskikh snimkakh (Application of the Technique of Multifractal Image Segmentation to Aerospace-Image Edge Detecting), Izvestiya Vuzov. Geodeziya i aerofotosemka, 2008, No. 6, pp. 37–41.
  2. Malinnikov V. A., Uchaev D. V., Uchaev Dm. V., Razrabotka metoda obobshchennogo lokal’no-global’nogo mul’tifraktal’nogo analiza izobrazhenii dlya issledovaniya prostranstvennoi struktury slozhnykh prirodno-antropogennykh sistem (A technique of generalized local-global multifractal analysis of images for research of spatial structure of complex natural-anthropogenic systems), Izvestiya Vuzov. Geodeziya i aerofotosemka, 2010, No. 4, pp. 64–68.
  3. Malinnikova O. N., Malinnikov V. A., Uchaev Dm. V., Uchaev D. V., Asimmetriya mul’tifraktal’nykh spektrov, opisyvayushchikh poverkhnostnuyu strukturu uglei vybrosoopasnykh i nevybrosoopasnykh plastov (Asymmetry of multifractal spectra describing the surface structure of coals of outburst-hazardous and outburst-nonhazardous beds), Deformirovanie i razrushenie materialov s defektami i dinamicheskie yavleniya v gornykh porodakh i vyrabotkakh (Deformation and Destruction of Materials with Defects and Dynamic Phenomena in Rocks and Working), Proc. 24th Intern. Scientific School, Alushta, Sept. 22–28, 2014, Simferopol: Tavricheskii natsional’nyi universitet, 2014, pp. 125–130.
  4. Uchaev D. V., Uchaev Dm. V., Esipov A. S., Filatova E. G., Fraktal’nyi podkhod k vyboru koeffitsienta szhatiya giperspektral’nykh izobrazhenii v metode 3D-SPIHT pri uslovii posleduyushchei klassifikatsii vosstanavlivaemykh izobrazhenii metodom opornykh vektorov (Fractal approach to the choice of the compression ratio of hyperspectral images in the 3D–SPIHT method under the condition of subsequent classification of the decompressed images by the support vector machine), Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2017, Vol. 14, No. 4, pp. 9–23.
  5. Chaban L. N., Malinnikov V. A., Uchaev D. V., Uchaev Dm. V., Metody otbora informativnykh kanalov pri tematicheskoi obrabotke giperspektral’nykh izobrazhenii (Methods of selection of informative band for thematic processing of hyper-spectral images), Izvestiya Vuzov. Geodeziya i aerofotosemka, 2014, No. 4, pp. 63–74.
  6. Eismann M. T., Hyperspectral Remote Sensing, Bellingham, Washington: SPIE Press, 2012, 748 p.
  7. Fauvel M., Spectral and spatial methods for the classification of urban remote sensing data: Doctoral Thesis, Grenoble, 2007, 189 p.
  8. Fauvel M., Chanussot J., Benediktsson J. A., A spatial-spectral kernel-based approach for the classification of remote-sensing images, Pattern Recognition, 2012, Vol. 45, No. 1, pp. 381–392.
  9. Ghamisi P., Benediktsson J. A., Ulfarsson M. O., Spectral-Spatial Classification of Hyperspectral Images Based on Hidden Markov Random Fields, IEEE Transactions on Geoscience and Remote Sensing, 2014, Vol. 52, No. 5, pp. 2565–2574.
  10. Jia X., Kuo B.-C., Crawford M. M., Feature Mining for Hyperspectral Image Classification, Proc. IEEE, 2013, Vol. 101, No. 3, pp. 676–697.
  11. Kuo B.-C., Landgrebe D. A., Nonparametric weighted feature extraction for classification, IEEE Transactions on Geoscience and Remote Sensing, 2004, Vol. 42, No. 5, pp. 1096–1105.
  12. Lee C., Landgrebe D. A., Feature extraction based on decision boundaries, IEEE Transactions on Pattern Analysis and Machine Intelligence, 1993, Vol. 15, No. 4, pp. 388–400.
  13. Levy Vehel J., Mignot P., Berroir J.-P., Multifractals, texture, and image analysis, Proc. CVPR’92, Champaign, 1992, pp. 661–664.
  14. Melgani F., Bruzzone L., Classification of Hyperspectral Remote Sensing Images With Support Vector Machines, IEEE Transactions on Geoscience and Remote Sensing, 2004, Vol. 42, No. 8, pp. 1778–1790.
  15. Soille P., Beyond self-duality in morphological image analysis, Image and Vision Computing, 2005, Vol. 23, No. 2, pp. 249–257.
  16. Tarabalka Y., Chanussot J., Benediktsson J. A., Segmentation and classification of hyperspectral images using watershed transformation, Pattern Recognition, 2010, Vol. 43, No. 7, pp. 2367–2379.
  17. Uchaev Dm. V., Uchaev D. V., Malinnikov V. A., Image contrast enhancement using Chebyshev wavelet moments, Proc. SPIE, Barcelona, 2015, Vol. 9875, p. 987512.
  18. Uchaev Dm. V., Uchaev D. V., Malinnikov V. A., Chebyshev-based technique for automated restoration of digital copies of faded photographic prints, J. Electronic Imaging, 2017, Vol. 26, No. 1, p. 011024.
  19. Vapnik V. N., Statistical Learning Theory, 1st ed., New York: Wiley-Interscience, 1998, 768 p.
  20. Ye Z., Fowler J. E., Bai L., Spatial-spectral hyperspectral classification using local binary patterns and Markov random fields, J. Applied Remote Sensing, 2017, Vol. 11, No. 3, p. 035002.
  21. Zhang R., Tian J., Li Z., Sun X., Jiang X., Spatial scaling and information fractal dimension of surface parameters used in quantitative remote sensing, Intern. J. Remote Sensing, 2008, Vol. 29, No. 17, pp. 5145–5159.