A New Gabor Filter Based Kernel for Texture Classification with SVM

  • Mahdi Sabri
  • Paul Fieguth
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3212)


The performance of Support Vector Machines (SVMs) is highly dependent on the choice of a kernel function suited to the problem at hand. In particular, the kernel implicitly performs a feature selection which is the most important stage in any texture classification algorithm. In this work a new Gabor filter based kernel for texture classification with SVMs is proposed. The proposed kernel function is based on a Gabor filter decomposition and exploiting linear predictive coding (LPC) in each subband, and exploiting a filter selection method to choose the best filters. The proposed texture classification method is evaluated using several texture samples, and compared with recently published methods. The comprehensive evaluation of the proposed method shows significant improvement in classification error rate.


Texture Classification Support Vector Machine Linear Predictive Coding Gabor Filters Segmentation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Marti, J., Batlle, J., Casals, A.: Model-based objects recognition in industrial environments. In: Proc. ICRA, IEEE, Los Alamitos (1997)Google Scholar
  2. 2.
    Horng, M.H., Sun, Y.N., Lin, X.Z.: Texture feature coding for classification of liver. Computerized Medical Imaging and Graphics 26, 33–42 (2002)CrossRefGoogle Scholar
  3. 3.
    Kim, J., Park, H.: Statistical texture features for detection of microcalcifications. IEEE Transaction on Medical Imaging 18, 231–238 (1999)zbMATHCrossRefGoogle Scholar
  4. 4.
    Scholkopf, B., Sung, K., Burges, C.J.C., Girosi, F., Niyogi, P., Pogio, T., Vapnik, V.: Comparing support vector machines with gaussian kernels to radial basis function classifiers. IEEE Transaction on Signal Processing 45, 2765–2785 (1997)CrossRefGoogle Scholar
  5. 5.
    Kim, K.I., Jung, K., Park, S.H., Kim, H.J.: Support vector machine for texture classification. IEEE Transaction on PAMI 24, 1542–1550 (2002)Google Scholar
  6. 6.
    Rabiner, L., Juang, B.H.: Fundamentals of speech recognition. Printce Hall, Englewood Cliffs (1993)Google Scholar
  7. 7.
    Vapnik, V.: The Nature of Statistical Learning Theory. Springer, Heidelberg (1995)zbMATHGoogle Scholar
  8. 8.
    Hubel, D.H., Wiesel, T.N.: Receptive fields and functional architecture in two nonstriate visual areas 18 and 19 of the cat. J. Neurophysiol. 28, 229–289 (1965)Google Scholar
  9. 9.
    Clausi, D.A., Jerningan, M.E.: Designing gabor filters for optimal texture seprability. Pattern Recognition Letters 33, 1835–1849 (2000)Google Scholar
  10. 10.
    Davy, M., Doncarli, C.: A new non-stationary test procedure for improved loud speaker fault detection. J. Audio Eng. Soc. 50, 458–469 (2002)Google Scholar
  11. 11.
    Duda, R., Hart, P., Stork, D.: Pattern Classification. Wiley, New York (2000)Google Scholar
  12. 12.
    Brodatz, P.: Textures Album for Artists and Designers. New york (1966) Google Scholar
  13. 13.
    MIT Vision and Modeling Group (1998) Google Scholar
  14. 14.
    Manian, V., Vasquez, R., Katiyar, P.: Texture classification using logical operators. IEEE Transaction on Image Processing 9, 1693–1703 (2000)zbMATHCrossRefMathSciNetGoogle Scholar
  15. 15.
    Arivazhagan, S., Ganesan, L.: Texture classification using wavelet transform. Pattern Recognition Letters 24, 1513–1521 (2003)zbMATHCrossRefGoogle Scholar
  16. 16.
    Randen, T., Husoy, J.H.: Filtering for texture classification: A comparative study. IEEE Trans. on Pattern Recognit. Machine Intell. 21, 291–310 (1999)CrossRefGoogle Scholar
  17. 17.
    Liu, X., Wang, D.: Texture classification using spectral histogram. IEEE Trans. Image Processing 12, 661–670Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Mahdi Sabri
    • 1
  • Paul Fieguth
    • 1
  1. 1.Department of Systems Design EngineeringUniversity of WaterlooWaterlooCanada

Personalised recommendations