Skip to main content

Advertisement

SpringerLink
Log in

Experimental comparison of support vector machines with random forests for hyperspectral image land cover classification

  • Published:
Journal of Earth System Science Aims and scope Submit manuscript

Abstract

The performances of regular support vector machines and random forests are experimentally compared for hyperspectral imaging land cover classification. Special characteristics of hyperspectral imaging dataset present diverse processing problems to be resolved under robust mathematical formalisms such as image classification. As a result, pixel purity index algorithm is used to obtain endmember spectral responses from Indiana pine hyperspectral image dataset. The generalized reduced gradient optimization algorithm is thereafter executed on the research data to estimate fractional abundances in the hyperspectral image and thereby obtain the numeric values for land cover classification. The Waikato environment for knowledge analysis (WEKA) data mining framework is selected as a tool to carry out the classification process by using support vector machines and random forests classifiers. Results show that performance of support vector machines is comparable to that of random forests. This study makes a positive contribution to the problem of land cover classification by exploring generalized reduced gradient method, support vector machines, and random forests to improve producer accuracy and overall classification accuracy. The performance comparison of these classifiers is valuable for a decision maker to consider tradeoffs in method accuracy versus method complexity.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Figure 1
Figure 2
Figure 3
Figure 4
Figure 5

Similar content being viewed by others

References

  • Abadie J and Carpentier J 1969 Generalization of the Wolfe reduced gradient method in the case of non-linear constraints; In: Optimization (ed.) Fletcher R (London: Academic Press), pp. 37–47.

  • Abe B T, Olugbara O O and Marwala T 2012 Hyperspectral image classification using random forest and neural network; Lecture Notes in Engineering and Computer Science: Proceedings of the World Congress on Engineering and Computer Science, WCECS 2012, 24–26 October, San Francisco, USA, pp. 522–527.

  • Adams J B, Sabol D E, Kapos V, Filho R A, Roberts D A, Smith M O and Gillespie A R 1995 Classification of multispectral images based on fractions of endmembers: Application to land-cover change in the Brazilian Amazon; Remote Sens. Environ. 52(2) 137–154.

  • Bateson C A and Curtiss B 1996 A method for manual endmember selection and spectral unmixing; Remote Sens. Environ. 55 229–243.

  • Bateson C A, Asner G P and Wessman C A 2000 Endmember bundles: A new approach to incorporating endmember variability into spectral mixture analysis; IEEE Trans. Geosci. Remote Sens. 38 1083–1094.

  • Bishop C M 2006 Pattern recognition and machine learning, 1st edn, Springer.

  • Boardman J 1993 Automating spectral unmixing of AVIRIS data using convex geometry concepts; In: Summaries of airborne earth science workshop, JPL Publication 93–26, pp. 111–114.

  • Boardman J W, Kruse F A and Green R O 1995 Mapping target signatures via partial unmixing of AVIRIS data; In: Summaries of the VI JPL Airborn Earth Science Workshop, pp. 23–26.

  • Bosch A, Zisserman A and Munoz X 2007 Image classification using random forests and ferns; In: Proc. ICCV 1–8.

  • Bowles J, Palmadesso P J, Antoniades J A, Baumback M M and Rickard L J 1995 Use of filter vectors in hyperspectral data analysis; Proc. SPIE 2553 148–157.

  • Breiman L 1996 Bagging predictors; Machine Learning 24(2) 123–140.

  • Breiman L 2001 Random forests; Machine Learning 45(1) 5–32.

  • Bruzzone L and Cossu R 2002 A multiple-cascade-classifier system for a robust and partially unsupervised updating of land-cover maps; IEEE Trans. Geosci. Remote Sens. 40(9) 1984–1996.

  • Camps-Valls G and Bruzzone L 2005 Kernel based methods for hyperspectral image classification; IEEE Trans. Geosci. Remote Sens. 43(6) 1351–1362.

  • Chakrabarty A, Choudhury O, Sarkar P, Paul A and Sarkar D 2012 Hyperspectral image classification incorporating bacterial foraging-optimized spectral weighting; Artificial Intelligence Res. 1(1) 63–83.

  • Chaudhry F, Wu C, Liu W, Chang C-I and Plaza A 2006 Pixel purity index-based algorithms for endmember extraction from hyperspectral imagery; In: Recent Advances in Hyperspectral Signal and Image Processing (ed.) Chang C-I, Trivandrum, India: Research Signpost 3, pp. 31–61.

  • Chen Y, Crawford M and Ghosh J 2007 Knowledge based stacking of hyperspectral data for land cover classification; Proceedings of the IEEE Symposium on Computational Intelligence and Data Mining, pp. 316–322.

  • 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) 593–600.

  • Congalton R G 1991 A review of assessing the accuracy of classifications of remotely sensed data; Remote Sens. Environ. 37(1) 35–46.

  • Congalton R G and Mead R A 1983 A quantitative method to test for consistency and correctness in photo-interpretation; Photogramm. Eng. Remote Sens. 49 69–74.

  • Congalton R G, Oderwald R G and Mead R A 1983 Assessing landsat classification accuracy using discrete multivariate statistical techniques; Photogram. Eng. Remote Sens. 49 1671–1678.

  • Cortes C and Vapnik V 1995 Support-Vector Networks; Machine Learning 20(3) 273–297.

  • Demir B and Ertürk S 2007 Hyperspectral image classification using relevance vector machines; IEEE Geosci. Remote Sens. Lett. 4(4) 586–590.

  • Dobigeon N, Moussaoui S, Coulon M, Tourneret J-Y and Hero A O 2009 Joint Bayesian endmember extraction and linear unmixing for hyperspectral imagery; IEEE Trans. Signal Process. 57(11) 4355–4368.

  • Du Q, Raksuntorn N, Younan N H and King R L 2008 End-member extraction for hyperspectral image analysis; Appl. Opt. 47 77–84.

  • Ellis J M 2001 Searching for oil seeps and oil-impacted soil with hyperspectral imagery; Earth Observation Magazine, pp. 25–28.

  • Garner S R 1995 WEKA: The Waikato environment for knowledge analysis; Proceedings of the New Zealand Computer Science Research Students Conference, pp. 57–64.

  • Gómez D and Montero J 2011 Determining the accuracy in image supervised classification problems; Proceedings of the 7th Conference of the European Society for Fuzzy Logic and Technology (EUSFLAT-2011) and âœles rencontres francophones sur la Logique Floue et ses Applicationsâ (LFA-2011), Aix-les-Bains, France, pp. 342–349.

  • Gong P and Zhang A 1999 Noise effect on linear spectral unmixing; Geographic Inform. Sci. 5(1) 52–57.

  • Gonzalez C, Resano J, Mozos D, Plaza A and Valencia D 2010 FPGA implementation of the pixel purity index algorithm for remotely sensed hyperspectral image analysis; EURASIP; J. Adv. Signal Process. 9,69,806 1–13.

  • Hall F G, Strebel D E, Nickeson J E and Goets S J 1991 Radiometric rectification: Towards a common radiometric response among multidate, multisensor images; Remote Sens. Environ. 35 11–27.

  • Heinz D C and Chang C 2001 Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery; IEEE Trans. Geosci. Remote Sens. 39(3) 529–545.

  • Iordache M-D, Bioucas-Dias J M and Plaza A 2011 Sparse unmixing of hyperspectral data; IEEE Trans. Geosci. Remote Sens. 49(6) 2014–2039.

  • Japkowicz N and Shah M 2011 Evaluating Learning Algorithms: A Classification Perspective; Cambridge University Press.

  • Karaska M A, Hugenin R L, Beacham J L, Wang M, Jenson J R and Kaufman R S 2004 AVIRIS measurements of chlorophyll, suspended minerals, dissolved organic carbon and turbidity in the Neuse River, North Calina; Photogramm. Eng. Remote Sens. 70(1) 125–133.

  • Kärdi T 2007 Remote sensing of urban areas: Linear spectral unmixing of landsat thematic mapping images acquired over Tartu (Estonia); Proc. Estonian Acad. Sci. Biol. Ecol. 56(1) 19–32.

  • Keshava N and Mustard J F 2002 Spectral unmixing; IEEE Signal Processing Magazine 19(1) 44–57.

  • Lacar F M, Lewis M M and Grierson I T 2001 Use of hyperspectral imagery for mapping grape varieties in the Barossa Valley, South Australia; Geosci. Remote Sens. Symp. IGARSS ’01 6 2875–2877.

  • Landgrebe D 1998 Multispectral data analysis: A signal theory perspective. School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN.

  • Landgrebe D and Biehl L 2001 An Introduction to MultiSpec. School of Electrical and Computer Engineering, Purdue University.

  • Lasdon L S, Fox R L and Ratner M W 1974 Nonlinear optimization using the generalized reduced gradient method; Revue française d’ automatique, d’ informatique et de recherché 3 73–103.

  • Lennon M, Mercier G and Hubert-Moy L 2002 Classification of hyperspectral images with nonlinear filtering and support vector machines; In: Proc. IEEE Trans. Geosci. Remote Sens. Symposium, Toronto, ON, Canada, 3 1670–1672.

  • Li J 2011 Remote sensing image information mining with HPC cluster and DryadLINQ; Proceedings of the 49th ACM Southeast Conference ACM SE’11, pp. 227–232.

  • Liu Y and Zhen Y F 2005 One-against-all multi-class SVM classification using reliability measures; Proc. IEEE Int. Joint Conf. Neural Networks 2 849–854.

  • Liu H, Fan Y, Deng X and Ji S 2009 Parallel processing architecture of remotely sensed image processing system based on cluster; Image and Signal Process. CISP ’09, 2nd Int. Congress, pp. 1–4.

  • Martinez P J, Pérez R M, Plaza A, Aguilar P L, Cantero M C and Plaza J 2006 Endmember extraction algorithms from hyperspectral images; Ann. Geophys. 49(1) 93–101.

  • Melgani F and Bruzzone L 2004 Classification of hyperspectral remote sensing images with support vector machines; IEEE Trans. Geosci. Remote Sens. 42(8) 1778–1790.

  • Mountrakis G, Im J and Ogole C 2011 Support vector machines in remote sensing: A review; ISPRS J. Photogramm. Remote Sens. 66 247–259.

  • Nascimento J M P and Bioucas-Dias J M 2005 Does independent component analysis play a role in unmixing hyperspectral data? IEEE Trans. Geosci. Remote Sens. 43(1) 175–187.

  • Pal M 2005 Random forest classifier for remote sensing classification; Int. J. Remote Sens. 26(1) 217–222.

  • Palsson F, Sveinsson J R, Benediktsson J A and Aanaes H 2012 Classification of Pansharpened Urban Satellite Images; Selected Topics in Appl. Earth Observations and Remote Sens. IEEE J. 5(1) 281–297.

  • Plaza A, Martínez P, Plaza J and Pérez R 2003 Spectral analysis of hyperspectral image data; Advances in Technique for Analysis of Remotely Sensed Data, IEEE Workshop, pp. 298–307.

  • Plaza A, Martinez P, Perez R and Plaza J 2004 A quantitative and comparative analysis of endmember extraction algorithms from hyperspectral data; IEEE Trans. Geosci. Remote Sens. 42(3) 650–663.

  • Plaza A, Plaza J and Cristo A 2008 Morphological feature extraction and spectral unmixing of hyperspectral images; IEEE Int. Symp. Signal Process. Information Technol., pp. 12–17.

  • Rodriguez-Galiano V F, Ghimire B, Rogan J, Chica-Olmo M and Rigol-Sanchez J P 2011 An assessment of the effectiveness of a random forest classifier for land cover classification; ISPRS J. Photogramm. Remote Sens. 67(9) 3–104.

  • Rosenfield G and Fitzpatrick-Lins K 1986 A coefficient of agreement as a measure of thematic classification accuracy; Photogramm. Eng. Remote Sens. 52 223–227.

  • Sanchez S, Martin G, Plaza A and Chang C 2010 GPU implementation of fully constrained linear spectral unmixing for remotely sensed hyperspectral data exploitation; Proceedings SPIE Satellite Data Compression, Communications, and Processing VI 7810 78100G-1–78100G-11.

  • Settle J J and Drake N A 1993 Linear mixing and the estimation of ground cover proportions; Int. J. Remote Sens. 14(6) 1159–1177.

  • Shao Y and Lunetta R S 2012 Comparison of support vector machine, neural network, and CART algorithms for the land-cover classification using limited training data points; ISPRS J. Photogramm. Remote Sens. 70(0) 78–87.

  • Story M and Congalton R G 1986 Accuracy assessment: A user’s perspective; Photogramm. Eng. Remote Sens. 52 397–399.

  • Su C-T and Lii G-R 1995 Reliability optimization design of distribution systems via multi-level hierarchical procedures and generalized reduced gradient method, energy management and power delivery; Proceedings of EMPD ’95, Int. Conference 1 180–185.

  • Su H, Sheng Y and Du P 2008 A new band selection algorithm for hyperspectral data based on fractal dimension; Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci. 37 Part B7 279–283.

  • Tarabalka Y, Chanussot J and Benediktsson J A 2009 Classification based marker selection for watershed transform of hyperspectral images; Geosci. Remote Sens. Symp. IEEE Int. IGARSS 3 III–105–III–108.

  • Theiler J, Lavenier D, Harvey N, Perkins S and Szymanski J 2000 Using blocks of skewers for faster computation of pixel purity index; Proc. SPIE Int. Conf. Optical Sci. Technol. 4132 61–71.

  • Udelhoven T, Waske B, Linden S and Heitz S 2009 Land-cover classification of hypertemporal data using ensemble systems; IEEE Trans. Geosci. Remote Sens., pp. III-1012–III-1015.

  • Ul Haq Q S, Shi L, Tao L and Yang S 2010 Hyperspectral data classification via sparse representation in homotopy; Inf. Sci. Eng. (ICISE) 2010, 2nd Int. Conf., pp. 3748–3752.

  • Van der Meer F D and Jia X 2012 Collinearity and orthogonality of endmembers in linear spectral unmixing; Int. J. Appl. Earth Observ. Geoinform. 18 491–503.

  • Winter M E 1999 N-FINDR: An algorithm for fast autonomous spectral endmember determination in hyperspectral data; Proc. SPIE Imaging Spectrometry V 3753 266–275.

  • Witten I H and Frank E 2005 Data mining: Practical machine learning tools and techniques, 2nd edn, Morgan Kaufmann, San Francisco, pp. 176–178.

  • Wolter P T, Johnston C A and Niemi G 2005 Mapping submergent aquatic vegetation in the US great lakes using quickbird satellite data; Int. J. Remote Sens. 26 5255–5274.

  • Xie Y, Sha Z and Yu M 2008 Remote sensing imagery in vegetation mapping: A review; J. Plant Ecol. 1 9–23.

  • Zhang B, Sun X, Gao L and Yang L 2011 Endmember extraction of hyperspectral remote sensing images based on the ant colony optimization (ACO) algorithm; IEEE Trans. Geosci. Remote Sens. 49(7) 2635–2646.

  • Zhou G, Wu B and Li M 2011 Improved accuracy assessment indices for object-based high resolution remotely sensed imagery classification; Image Analysis and Signal Processing (IASP), 2011 International Conference 181–186 21–23.

Download references

Acknowledgements

The authors would like to thank Prof. D Landgrebe of the Department of Electrical and Computer Engineering and L Biehl both at the Purdue University for their MultiSpec software for the analysis of hyperspectral data (Landgrebe and Biehl 2001). In particular, authors acknowledge the professional assistance received from Prof. Landgrebe who approved the use of Indiana pines dataset for experimentation. Thanks for the support and the research grants from Witwatersrand University, Tshwane University of Technology and University of Johannesburg in South Africa.

Author information

Authors and Affiliations

  1. School of Electrical and Information Engineering, University of Witwatersrand, Johannesburg, South Africa.

    B T Abe

  2. Department of Information Technology, Durban University of Technology, Durban, South Africa.

    O O Olugbara

  3. Department of Engineering and the Built Environment, University of Johannesburg, Johannesburg, South Africa.

    T Marwala

Authors
  1. B T Abe
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. O O Olugbara
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. T Marwala
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to B T Abe.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Abe, B.T., Olugbara, O.O. & Marwala, T. Experimental comparison of support vector machines with random forests for hyperspectral image land cover classification. J Earth Syst Sci 123, 779–790 (2014). https://doi.org/10.1007/s12040-014-0436-x

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12040-014-0436-x

Keywords

Search

Navigation