Classification of Hyperspectral Images Using Machine Learning Methods

  • Bolanle Tolulope Abe
  • Oludayo O. Olugbara
  • Tshilidzi Marwala
Chapter
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 247)

Abstract

Mixed pixels problem has significant effects on the application of remote sensing images. Spectral unmixing analysis has been extensively used to solve mixed pixels in hyperspectral images. This is based on the knowledge of a set of unidentified endmembers. This study used pixel purity index to extract endmembers from hyperspectral dataset of Washington DC mall. Generalized reduced gradient (GRG) a mathematical optimization method is used to estimate fractional abundances (FA) in the dataset. WEKA data mining tool is chosen to develop ensemble and non-ensemble classifiers using the set of the FA. Random forest (RF) and bagging represent ensemble methods while neural networks and C4.5 represent non-ensemble models for land cover classification (LCC). Experimental comparison between the classifiers shows that RF outperforms all other classifiers. The study resolves the problem associated with LCC by using GRG algorithm with supervised classifiers to improve overall classification accuracy. The accuracy comparison of the learners is important for decision makers in order to consider tradeoffs in accuracy and complexity of methods.

Keywords

Accuracy Classifier Ensemble Hyperspectral Image Optimization 

Notes

Acknowledgments

This work was supported by University of Johannesburg, South Africa, Tshwane University of Technology, South Africa and University of the Witwatersrand, Johannesburg. South Africa.

References

  1. 1.
    Xie Y, Sha Z, Yu M (2008) Remote sensing imagery in vegetation mapping: a review. J Plant Ecol 1:9–23CrossRefGoogle Scholar
  2. 2.
    Shaw GA, Burke HK (2003) Spectral imaging for remote sensing. Lincoln Lab J 14(1):3–28Google Scholar
  3. 3.
    Chang CI, Heinz DC (2000) Constrained subpixel target detection for remotely sensed imagery. IEEE Trans Geosci Remote Sens 38(3):1144–1159CrossRefGoogle Scholar
  4. 4.
    Abe BT, Olugbara OO, Marwala T (2012) Hyperspectral image classification using random forest and neural network. In: Proceedings of the world congress on engineering and computer science 2012, Lecture notes in engineering and computer science, 24–26 October, San Francisco, USA, pp 522–527Google Scholar
  5. 5.
    Sanchez S, Martin G, Plaza A, Chang C (2010) GPU implementation of fully constrained linear spectral unmixing for remotely sensed hyperspectral data exploitation. In: Proceedings SPIE satellite data compression, communications, and processing VI, 2010, vol 7810, pp 78100G-1–78100G-11Google Scholar
  6. 6.
    Iordache M-D, Bioucas-Dias JM, Plaza A (2011) Sparse unmixing of hyperspectral data. Geosci Remote Sens IEEE Trans 49(6):2014–2039CrossRefGoogle Scholar
  7. 7.
    Zhang B, Sun X, Gao L, Yang L (2011) Endmember extraction of hyperspectral remote Sensing images based on the ant colony optimization (ACO) algorithm. Geosci Remote Sens IEEE Trans 49(7):2635–2646CrossRefGoogle Scholar
  8. 8.
    Martinez PJ, Perez RM, Plaza A, Aguilar PL, Cantero MC, Plaza J (2006) Endmember extraction algorithms for hyperspectral images. Ann Geophy 49(1):93–101Google Scholar
  9. 9.
    Landgrebe DA (2003) Signal theory methods in multispectral remote sensing. Wiley, HobokenCrossRefGoogle Scholar
  10. 10.
    Abadie J, Carpentier J (1969) Generalization of the Wolfe reduced gradient method in the case of non-linear constraints. In: Fletcher R (ed) Optimization. Academic Press, London, pp 37–47Google Scholar
  11. 11.
    Lasdon LS, Fox RL, Ratner MW (1974) Nonlinear optimization using the generalized reduced gradient method. Revue française d’ automatique, d’ informatique et de recherché, 1974, issue 3, pp 73–103Google Scholar
  12. 12.
    Maree R, Stevens B, Geurts P, Guern Y, Mack P (2009) A machine learning approach for material detection in hyperspectral images. In: Proceedings of IEEE computer society conference on computer vision and pattern recognition workshops, 2009, pp 106–111Google Scholar
  13. 13.
    Okori W, Obua J (2011) Machine learning classification technique for famine prediction. In: Proceedings of the world congress on engineering, 2011, vol II, July 6–8 2011, London, U.K, pp 991–996Google Scholar
  14. 14.
    Abe B, Gidudu A, Marwala T (2010) Investigating the effects of ensemble classification on remotely sensed data for land cover mapping. In: Proceedings of IEEE international conference on geoscience and remote sensing symposium (IGARSS), 2010, pp 2832–2835Google Scholar
  15. 15.
    Govindarajan M, Chandrasekaran RM (2012) Intrusion detection using ensemble of classification methods. In: Proceedings of the world congress on engineering and computer science 2012, Lecture notes in engineering and computer science, 24–26 October 2012, San Francisco, USA, pp 459–464Google Scholar
  16. 16.
    Benediktsson JA, Swain PH, Ersoy OK (1990) Neural network approaches versus statistical methods in classification of multisource remote sensing data. IEEE Trans Geosci Remote Sens 28:540–552CrossRefGoogle Scholar
  17. 17.
    Ramírez-Quintana JA, Chacon-Murguia MI, Chacon-Hinojos JF (2012) Artificial neural image processing applications: a survey. Eng Lett 20:1–68 ([Online] Available http://www.engineeringletters.com/issues_v20/issue_1/EL_20_1_09.pdf)Google Scholar
  18. 18.
    Mekanik F, Imteaz MA (2012) A multivariate artificial neural network approach for rainfall forecasting: case study of Victoria, Australia. In: Proceedings of the world congress on engineering and computer science 2012, Lecture notes in engineering and computer science, 24–26 October 2012, San Francisco, USA, pp 557–561Google Scholar
  19. 19.
    Wang P, Zhang J, Jia W, Lin Z (2008) A study on decision tree classification method of land use/land cover -taking tree counties in Hebei province as an example. In: Proceedings of earth observation and remote sensing applications, 2008, international workshop on 2008, pp 1–5Google Scholar
  20. 20.
    Pinho CMD, Silva FC, Fonseca LMC, Monteiro AMV (2008) Intra-urban land cover classification from high-resolution images using the C4.5 algorithm. In: Proceedings of the international archives of the photogrammetry, remote sensing and spatial information sciences, 2008, vol 37(B7), pp 695–699Google Scholar
  21. 21.
    Breiman L (1996) Bagging predictors. Mach Learn 24(2):123–140MathSciNetMATHGoogle Scholar
  22. 22.
    Breiman L (2001) Random forests. Mach Learn 45(1):5–32CrossRefMATHGoogle Scholar
  23. 23.
    Kärdi T (2007) Remote sensing of sensing of urban areas: linear spectral unmixing of landsat thematic Mapping images acquired over Tartu (Estonia). In: Proceedings of the Estonian academy of sciences: biology, ecology, 2007, vol 56, no 1, pp 19–32Google Scholar
  24. 24.
    Theiler J, Lavenier D, Harvey N, Perkins S, Szymanski J (2000) Using blocks of skewers for faster computation of pixel purity index. In: Proceedings of the SPIE international conference on optical science and technology, 2000, no 4132, pp 61–71Google Scholar
  25. 25.
    Keshava N, Mustard JF (2002) Spectral unmixing. IEEE Signal Process Mag 19(1):44–57CrossRefGoogle Scholar
  26. 26.
    Chaudhry F, Wu C, Liu W, Chang C-I, Plaza A (2006) Pixel purity index-based algorithms for endmember extraction from hyperspectral imagery. In: Proceedings of recent advances in hyperspectral signal and image processing, C.-I Chang, Ed. Trivandrum, India: Research Signpost, 2006, no 3, pp 31–61Google Scholar
  27. 27.
    Plaza A, Martinez P, Perez R, Plaza J (2004) A quantitative and comparative analysis of endmember extraction algorithms from hyperspectral data. IEEE Trans Geosci Remote Sens 42(3):650–663CrossRefGoogle Scholar
  28. 28.
    Gonzalez C, Resano J, Mozos D, Plaza A, Valencia D (2010) FPGA implementation of the pixel purity index algorithm for remotely sensed hyperspectral image analysis. EURASIP J Adv Signal Process 969806:1–13Google Scholar
  29. 29.
    Boardman JW, Biehl LL, Clark RN, Kruse FA, Mazer AS, Torson J (2006) Development and implementation of software systems for imaging spectroscopy. In: Proceedings of IEEE international conference on geoscience and remote sensing symposium, 2006, July 31 2006–Aug 4 2006, pp 1969–1973Google Scholar
  30. 30.
    Garner SR (1995) WEKA: the Waikato environment for knowledge analysis. In: Proceedings of the NewZealand computer science research students conference, 1995, pp 57–64Google Scholar
  31. 31.
    Benediktsson JA, Swain PH, Ersoy OK (1990) Neural network approaches versus statistical methods in classification of multisource remote sensing data. IEEE Trans Geosci Remote Sens 28:540–552CrossRefGoogle Scholar
  32. 32.
    Congalton RG (1991) A review of assessing the accuracy of classifications of remotely sensed data. Remote Sens Environ 37(1):35–46CrossRefGoogle Scholar
  33. 33.
    Witten IH, Frank E (2005) Data mining: practical machine learning tools and techniques, 2nd edn. Morgan Kaufmann, San Francisco, pp 189–199, 316–319, 337–423Google Scholar
  34. 34.
    Congalton R (1988) A comparison of sampling schemes used in generating error matrices for assessing the accuracy of maps generated from remotely sensed data. Photogram Eng Remote Sens 54(5):593–600Google Scholar
  35. 35.
    Kumar Y, Sahoo G (2012) Analysis of parametric and non parametric classifiers for classification technique using WEKA. Int J Inf Technol Comput Sci 7:43–49. doi: 10.5815/ijitcs.2012.07.06 (Published Online July 2012 in MECS http://www.mecs-press.org/)
  36. 36.
    Hall M, Frank E, Holmes G, Pfahringer B, Reutemann P, Witten I (2009) The WEKA data mining software: an update. ACM SIGKDD explorations newsletter, 2009, vol 11Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  • Bolanle Tolulope Abe
    • 1
    • 2
  • Oludayo O. Olugbara
    • 3
  • Tshilidzi Marwala
    • 4
  1. 1.School of Electrical and Information EngineeringUniversity of the WitwatersrandJohannesburgSouth Africa
  2. 2.Department of Electrical EngineeringTshwane University of TechnologyPretoriaSouth Africa
  3. 3.Depatment of Information TechnologyDurban University of TechnologyDurbanSouth Africa
  4. 4.University of JohannesburgJohannesburgSouth Africa

Personalised recommendations