Spectral-spatial classification with k-means++ particional clustering
E.A. Zimichev, N.L. Kazanskiy, P.G. Serafimovich

PDF, 458 kB

Full text of article: Russian language.

DOI: 10.18287/0134-2452-2014-38-2-281-286

Pages: 281-286.

Abstract:
A complex spectral–spatial classification scheme for hyperspectral images is proposed and explored. The key feature of method is using widespread and simple enough algorithms while having high precision. The method combines the results of a pixel wise support vector machine classification and the segmentation map obtained by partitional clustering using majority voting. The k-means++ clusterization algorithm is used for image clustering. Principal component analysis is used to prevent redundant processing of similar data. The proposed method provides improved precision and speed of hyperspectral data classification.

Key words:
hyperspectral imaging, classification, segmentation, SVM, k-means.

References:

  1. Fursov, V.A. Thematic classification of hyperspectral images by conjugacy indicator / V.A. Fursov, S.A. Bibikov, O.A. Bajda // Computer Optics. – 2014. – V. 38, Issue 1. – P. 154-158.
  2. Zhuravel, Yu.N. The features of hyperspectral remote sensing data processing under environment monitoring task solution / Yu.N. Zhuravel, A.A. Fedoseev // Computer Optics. – 2013. – V. 37(4). – P. 471-476.
  3. Gashnikov, М.V. Hierarchical grid interpolation for hyperspectral image compression / М.V. Gashnikov, N.I. Glumov // Computer Optics. – 2014. – V. 38(1). – P. 87-93.
  4. Green, R.O. Imaging spectroscopy and the airborne visible/infrared imaging spectrometer (AVIRIS) / R.O. Green [et al.] // Remote Sensing of Environment. – 1998. – V. 65(3). – P. 227-248.
  5. Rickard, L.J. HYDICE: An airborne system for hyperspectral imaging / L.J. Rickard [et al.] // Optical Engineering and Photonics in Aerospace Sensing. – International Society for Optics and Photonics, 1993. – P. 173-179.
  6. Cristianini, N. An introduction to support vector machines and other kernel-based learning methods / N. Cristianini, J. Shawe-Taylor. – Cambridge university press, 2000.
  7. Gualtieri, J.A. Support vector machines for hyperspectral remote sensing classification / J.A. Gualtieri, R.F. Cromp // The 27th AIPR Workshop: Advances in Computer-Assisted Recognition. – International Society for Optics and Photonics, 1999. – P. 221-232.
  8. Tarabalka, Y. Spectral–spatial classification of hyperspectral imagery based on partitional clustering techniques / Y. Tarabalka, J.A. Benediktsson, J. Chanussot // IEEE Transactions on Geoscience and Remote Sensing. – 2009. – V. 47(8). – P. 2973-2987.
  9. Ball, G.H. ISODATA, a novel method of data analysis and pattern classification / G.H. Ball, D.J. Hall. – Stanford research institute Publisher, 1965.
  10. Dempster, A.P. Maximum likelihood from incomplete data via the EM algorithm / A.P. Dempster [et al.] //Journal of the Royal Statistical Society. – 1977. – V. 39(1). – P. 1-38.
  11. Vapnik, V. The nature of statistical learning theory. – Springer, 2000.
  12. Fauvel, M. Spectral and spatial methods for the classification of urban remote sensing data // Institut Technologique de Grenoble–Université d’Islande, Thèse de Doctorat, 2007.
  13. Rodarmel, C. Principal component analysis for hyperspectral image classification / C. Rodarmel, J. Shan // Surveying and Land Information Science. – 2002. – V. 62(2). – P. 115-122.
  14. Arthur, D. k-means++: The advantages of careful seeding / D. Arthur, S. Vassilvitskii // Proceedings of the eighteenth annual ACM-SIAM symposium on discrete algorithms. – Society for Industrial and Applied Mathematics, 2007. – P. 1027-1035.
  15. He, L. Fast connected-component labeling / L. He [et al.] // Pattern Recognition. – 2009. – V. 42(9). – P. 1977-1987.
  16. Bahmani, B. Scalable k-means++ / B. Bahmani [et al.] // Proceedings of the VLDB Endowment. – 2012. – V. 5(7). – P. 622-633.
  17. Scholkopf, B. Kernel principal component analysis / B. Scholkopf, A. Smola, K.R. Müller //Advances in kernel methods-support vector learning. – MIT Press Cambridge. – 1999. – P. 327-352.
© 2009, IPSI RAS
151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20