Advertisement

Image Classification into Object / Non-object Classes

  • Sungyoung Kim
  • Sojung Park
  • Minhwan Kim
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3115)

Abstract

We propose a method that automatically classifies the images into the object and non-object images. An object image is an image with object(s). An object in an image is defined as a set of regions located near the center of the image, which has significant color distribution compared with its surrounding (or background) region. We define three measures for the classification based on the characteristics of an object. The center significance is calculated from the difference in color distribution between the center area and its surrounding region. Second measure is the variance of significantly correlated colors in the image plane. Significantly correlated colors are first defined as the colors of two adjacent pixels that appear more frequently around center of an image rather than at the background of the image. The last one is the edge strength at the boundary of the region that is estimated as an object. To classify the images we combine each measure by training the neural network. A test with 900 images shows a classification accuracy of 84.2%. We also compare our result with the performance of several other classifiers, Naïve Bayes, Decision Table, and Decision Tree.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Vailaya, A., Jain, A.K., Zhang, H.J.: On Image Classification: City images vs. landscape. Pattern Recognition 31(12), 1921–1936 (1998)CrossRefGoogle Scholar
  2. 2.
    Szummer, M., Picard, R.W.: Indoor-outdoor image classification. In: IEEE Int’l Workshop Content-Based Access Image Video Databases, pp. 42–51 (1998)Google Scholar
  3. 3.
    Vailaya, A., Figueiredo, M.A.T., Jain, A.K., Zhang, H.J.: Image Classification for Content-Based Indexing. IEEE Trans. on Image Processing 10(1), 117–130 (2001)zbMATHCrossRefGoogle Scholar
  4. 4.
    Kim, S.Y., Park, S.Y., Kim, M.H.: In: Bakker, E.M., Lew, M., Huang, T.S., Sebe, N., Zhou, X.S. (eds.) CIVR 2003. LNCS, vol. 2728, pp. 39–49. Springer, Heidelberg (2003)CrossRefGoogle Scholar
  5. 5.
    Serra, J.R., Subirana, J.B.: Texture Frame Curves and Regions of Attention Using Adaptive Non-cartesian Networks. Pattern Recognition 32, 503–515 (1999)CrossRefGoogle Scholar
  6. 6.
    Park, S.B., Lee, J.W., Kim, S.K.: Content-based image classification using a neural network. Pattern Recognition Letter 25, 287–300 (2004)CrossRefGoogle Scholar
  7. 7.
    Huang, J., Kumar, S.R., Mitra, M., Zhu, W.J., Zabih, R.: Image Indexing Using Color Correlograms. In: Proc. Computer Vision and Pattern Recognition, pp. 762–768 (1997)Google Scholar
  8. 8.
    Lippmann, R.P.: An introduction to computing with neural nets. IEEE ASSP Magazine, 4–22 (1994)Google Scholar
  9. 9.
    Witten, I.H., Frank, E.: Data Mining. Academic Press, London (2000)Google Scholar
  10. 10.
    Good, I.J.: The estimation of Probabilities: An Essay on Modern Bayesian Methods. MIT Press, MA (1965)Google Scholar
  11. 11.
    Kohavi, R.: The Power of Decision Tables. In: Lavrač, N., Wrobel, S. (eds.) ECML 1995. LNCS, vol. 912, Springer, Heidelberg (1995)Google Scholar
  12. 12.
    Breiman, L., Friedman, J.H., Olshen, R.A., and Stone, C.J.: Classification and Regression Trees. Wadsworth, Belmont (1984)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Sungyoung Kim
    • 1
  • Sojung Park
    • 2
  • Minhwan Kim
    • 2
  1. 1.School of Computer EngineeringKumoh National Institute of TechnologyGumiKorea
  2. 2.Dept. of Computer EngineeringPusan National Univ.PusanKorea

Personalised recommendations