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


Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2017, Vol. 14, No. 4, pp. 9-23

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

D.V. Uchaev 1 , Dm.V. Uchaev 1 , A.S. Esipov 1 , E.G. Filatova 1 
1 Moscow State University of Geodesy and Cartography, Moscow, Russia
Accepted: 25.05.2017
DOI: 10.21046/2070-7401-2017-14-4-9-23
During the past decade, there has been a considerable growth of interest in the solution of applied problems using hyperspectral remote sensing images. Along with this growth, the need for highly effective methods of hyperspectral image compression also increases. 3D-SPIHT methods are one of the most effective methods used to compress hyperspectral images. One of the most important problems in the use of 3D-SPIHT methods is the choice of the compression ratio in conditions when decompressed hyperspectral images should be further subjected to classification. In this paper, results of a research on the impact of 3D-SPIHT compression of hyperspectral images on the quality of their classification by the support vector machine are presented. Using two test data sets, “Pavia University” (ROSIS) and “Salinas” (AVIRIS), it is shown that there is a relationship between the quality of classification and full-reference quality metrics of decompressed hyperspectral images. This relationship allows distinguishing three typical ranges of the compression ratio, at which the overall accuracy of the classification does not practically decrease, decreases slightly and decreases greatly. Based on the observations, a 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 is proposed.
Keywords: hyperspectral image, 3D-SPIHT, support vector machine, classification quality, fractal
Full text


  1. Vapnik V.N., Chervonenkis A.Ya., Teoriya raspoznavaniya obrazov. Statisticheskie problemy obucheniya (Theory of pattern recognition. Statistical problems of training), Moscow: Nauka, 1974, 416 p.
  2. V’yugin V.V., Matematicheskie osnovy mashinnogo obucheniya i prognozirovaniya (Mathematical Foundations of Machine Learning and Forecasting), Moscow: MTsNMO, 2014, 304 p.
  3. Selomon D., Szhatie dannykh, izobrazhenii i zvuka (A Guide to Data Compression Methods), Moscow: Tekhnosfera, 2004, 368 p.
  4. Uchaev D.V., Bobkov A.E., Malinnikov V.A., Uchaev Dm.V., Upravlenie giperspektral’nymi izobrazheniyami v protsesse nauchno-issledovatel’skoi deyatel’nosti malykh kollektivov (Management of hyperspectral images for scientific research of small teams), Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2016, Vol. 13, No. 6, pp. 233–248.
  5. Abousleman G.P., Marcellin M.W., Hunt B.R. Compression of hyperspectral imagery using the 3-D DCT and hybrid DPCM/DCT, IEEE Trans. Geosci. Remote Sens., 1995, Vol. 33, No. 1, pp. 26–34.
  6. Bioucas-Dias J.M., Plaza A., Camps-Valls G., Scheunders P., Nasrabadi N., Chanussot J., Hyperspectral Remote Sensing Data Analysis and Future Challenges, IEEE Geosci. Remote Sens. Mag., 2013, Vol. 1, No. 2, pp. 6–36.
  7. Boser B.E., Guyon I.M., Vapnik V.N., A Training Algorithm for Optimal Margin Classifiers, Proceedings of the Fifth Annual Workshop on Computational Learning Theory (COLT’92), New York, 1992, pp. 144–152.
  8. Chang C.-I., Hyperspectral data processing. Algorithm design and analysis, New York: Wiley, 2013, 1151 p.
  9. Christophe E., Mailhes C., Duhamel P., Best Anisotropic 3-D Wavelet Decomposition in a Rate-Distortion Sense, Proc. of the IEEE International Conference on Acoustics Speech and Signal Processing, 2006 (ICASSP 2006), Toulouse, 2006, pp. II-17-II-20.
  10. Congalton R.G., Green K., Assessing the Accuracy of Remotely Sensed Data: Principles and Practices, Boca Raton: CRC Press, 2008, 210 p.
  11. Cover T.M., Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition, IEEE Trans. Electron. Comput., 1965, Vol. EC–14, No. 3, pp. 326–334.
  12. Dragotti P.L., Poggi G., Ragozini A.R.P., Compression of multispectral images by three–dimensional SPIHT algorithm, IEEE Trans. Geosci. Remote Sens., 2000, Vol. 38, No. 1, pp. 416–428.
  13. Gamba P., A collection of data for urban area characterization, Proc. of the IEEE International Geoscience and Remote Sensing Symposium, 2004 (IGARSS 2004), Anchorage, 2004, pp. 69–72.
  14. García-Vílchez F., Muñoz-Marí J., Zortea M., Blanes I., González–Ruiz V., Camps-Valls G., Plaza A., Serra-Sagristà J., On the Impact of Lossy Compression on Hyperspectral Image Classification and Unmixing, IEEE Geosci. Remote Sens. Lett., 2011, Vol. 8, No. 2, pp. 253–257.
  15. Gualtieri J.A., Chettri S.R., Cromp R.F., Johnson L.F., Support Vector Machine Classifiers as Applied to AVIRIS Data, Summaries of the Eighth JPL Airborne Earth Science Workshop, Nevada, 1999, pp. 217–227.
  16. Kim B.-J., Pearlman W.A., An embedded wavelet video coder using three-dimensional set partitioning in hierarchical trees (SPIHT), Proc. of the Data Compression Conference, 1997 (DCC’97), Snowbird, 1997, pp. 251–260.
  17. Kim B.-J., Xiong Z., Pearlman W.A., Low bit-rate scalable video coding with 3-D set partitioning in hierarchical trees (3-D SPIHT), IEEE Trans. Circuits Syst. Video Technol., 2000, Vol. 10, No. 8, pp. 1374–1387.
  18. Langevin Y., Forni O., Image and spectral image compression for four experiments on the ROSETTA and Mars Express missions of ESA, Proc. SPIE. Applications of Digital Image Processing XXIII, 2000, Vol. 4115, pp. 364–373.
  19. Lee C., Choi E., Jeong T., Lee S., Lee J., Compression of hyperspectral images with discriminant features enhanced, J. Appl. Remote Sens., 2010, Vol. 4, No. 1, pp. 041764.
  20. Lee C., Youn S., Baek J.Y., Sagristà J.S., Effects of compression on classification performance and discriminant information preservation in remotely sensed data, Proc. SPIE. Satellite Data Compression, Communications, and Processing XI, 2015, Vol. 9501, pp. 950103.
  21. Liang Z., Xinming T., Guo Z., Xiaoliang W. Effects of JPEG2000 and SPIHT Compression on Image Classification, Proc. of the XXIst ISPRS Congress Technical Commission VII, 2008, pp. 541–544.
  22. Lim S., Sohn K., Lee C., Compression for hyperspectral images using three dimensional wavelet transform, Proc. of the IEEE International Geoscience and Remote Sensing Symposium, 2001 (IGARSS 2001). Scanning the Present and Resolving the Future, Sydney, 2001, pp. 109–111.
  23. Liu X., Beltran J.F., Mohanchandra N., Toussaint G.T., On Speeding Up Support Vector Machines: Proximity Graphs Versus Random Sampling for Pre-Selection Condensation, Intern. J. Computer, Electrical, Automation, Control and Information Engineering, 2013, Vol. 7, No. 1, pp. 133–140.
  24. Lossless Multispectral and Hyperspectral Image Compression standard. Green Book. Informational Report, Issue 1. CCSDS 120.2-G-1, Washington: The Consultative Committee for Space Data Systems, 2015, 99 p.
  25. Markman D., Malah D., Hyperspectral image coding using 3D transforms, Proc. of the International Conference on Image Processing, 2001 (ICIP 2001), Thessaloniki, 2001, Vol. 1, pp. 114–117.
  26. Melgani F., Bruzzone L., Classification of Hyperspectral Remote Sensing Images With Support Vector Machines, IEEE Trans. Geosci. Remote Sens., 2004, Vol. 42, No. 8, pp. 1778–1790.
  27. Pearlman W.A., Said A., Data compression using set partitioning in hierarchical trees, Patent, 5764807 United States, 1998.
  28. Penna B., Tillo T., Magli E., Olmo G., Transform Coding Techniques for Lossy Hyperspectral Data Compression, IEEE Trans. Geosci. Remote Sens., 2007, Vol. 45, No. 5, pp. 1408–1421.
  29. Qian S.-E., Hollinger A.B., Dutkiewicz M., Tsang H., Zwick H., Freemantle J.R., Effect of lossy vector quantization hyperspectral data compression on retrieval of red-edge indices, IEEE Trans. Geosci. Remote Sens., 2001, Vol. 39, No. 7, pp. 1459–1470.
  30. Said A., Pearlman W.A., A new, fast, and efficient image codec based on set partitioning in hierarchical trees, IEEE Trans. Circuits Syst. Video Technol., 1996, Vol. 6, No. 3, pp. 243–250.
  31. Serra-Sagristà J., Aulí-Llinàs F., Remote Sensing Data Compression, Computational Intelligence for Remote Sensing Studies in Computational Intelligence, Berlin, Heidelberg: Springer, 2008, pp. 27–61.
  32. Shah C.A., Watanachaturaporn P., Varshney P.K., Arora M.K., Some recent results on hyperspectral image classification, Proc. of the IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003 (WARSD 2003), Greenbelt, 2003, pp. 346–353.
  33. Tang X., Cho S., Pearlman W.A., 3D set partitioning coding methods in hyperspectral image compression, Proc. of the International Conference on Image Processing, 2003 (ICIP 2003), Barcelona, 2003, Vol. 2, pp. II-239–II-242.
  34. Tang X., Pearlman W.A., Lossy-To-Lossless Block-Based Compression of Hyperspectral Volumetric Data, Proc. of the International Conference on Image Processing, 2006 (ICIP 2006), Atlanta, 2006, pp. 1133–1136.
  35. Vaiopoulos A.D., Developing Matlab scripts for image analysis and quality assessment, Proc. SPIE. Earth Resources and Environmental Remote Sensing, GIS Applications II, 2011, Vol. 8181, pp. 81810B.
  36. van der Linden S., Rabe A., Held M., Jakimow B., Leitão P.J., Okujeni A., Schwieder M., Suess S., Hostert P., The EnMAP-Box — A Toolbox and Application Programming Interface for EnMAP Data Processing, Remote Sens., 2015, Vol. 7, No. 9, pp. 11249–11266.
  37. Vapnik V.N., The Nature of Statistical Learning Theory, New York: Springer Science & Business Media, 1995, 201 p.