This paper describes a feature extraction method for realtime surveillance. Eigenspace models are a convenient way to represent set of images with widespread applications. In the traditional approach to calculate these eigenspace models, known as batch PCA method, model must capture all the images needed to build the internal representation. This approach has some drawbacks. Since the entire set of images is necessary, it is impossible to make the model build an internal representation while exploring a new person. Updating of the existing eigenspace is only possible when all the images must be kept in order to update the eigenspace, requiring a lot of storage capability. In this paper we propose a method that allows for incremental eigenspace update method by incremental kernel PCA for realtime surveillance. Experimental results indicate that accuracy of proposed method is comparable to batch KPCA and outperforms than APEX. Furthermore proposed method has efficiency in memory requirement compared to KPCA.


Face Recognition Reconstruction Error Kernel Matrix Time Surveillance Apex Model 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Chellappa, R., Wilson, C.L., Sirohey, S.: Human and machine recognition of faces: a survey. Proc. of IEEE 83(5), 705–740 (1995)CrossRefGoogle Scholar
  2. 2.
    Brunelli, R., Poggio, T.: Face recognition: feature versus templates. IEEE Trans. PAMI 15(10), 1042–1052 (1993)Google Scholar
  3. 3.
    Turk, M., Pentland, A.: Face recognition using eigenfaces. In: Proc. IEEE Conf. on CVPR, pp. 586–591 (1991)Google Scholar
  4. 4.
    Winkeler, J., Manjunath, B.S., Chandrasekaran, S.: Subset selection for active object recognition. In: CVPR, vol. 2, pp. 511–516. IEEE Computer Society Press, Los Alamitos (1999)Google Scholar
  5. 5.
    Hall, P., Marshall, D., Martin, R.: On-line eigenalysis for classification. In: British Machine Vision Conference, vol. 1, pp. 286–295 (September 1998)Google Scholar
  6. 6.
    Softky, W.S., Kammen, D.M.: Correlation in high dimensional or asymmetric data set: Hebbian neuronal processing. Neural Networks 4, 337–348 (1991)CrossRefGoogle Scholar
  7. 7.
    Gupta, H., Agrawal, A.K., Pruthi, T., Shekhar, C., Chellappa, R.: An Experimental Evaluation of Linear and Kernel-Based Methods for Face Recognition, accessible at
  8. 8.
    Murakami, H., Kumar, B.V.K.V.: Efficient calculation of primary images from a set of images. IEEE PAMI 4(5), 511–515 (1982)Google Scholar
  9. 9.
    Vapnik, V.N.: Statistical learning theory. John Wiley & Sons, New York (1998)MATHGoogle Scholar
  10. 10.
    Scholkopf, B., Smola, A., Muller, K.R.: Nonlinear component analysis as a kernel eigenvalue problem. Neural Computation 10(5), 1299–1319 (1998)CrossRefGoogle Scholar
  11. 11.
    Tsuda, K.: Support vector classifier based on asymmetric kernel function. In: Proc. ESANN (1999)Google Scholar
  12. 12.
    Mika, S.: Kernel algorithms for nonlinear signal processing in feature spaces. Master’s thesis, Technical University of Berlin (November 1998)Google Scholar
  13. 13.
    Diamantaras, K.I., Kung, S.Y.: Principal Component Neural Networks: Theory and Applications. John Wiley & Sons, Inc., New York (1996)MATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Byung-Joo Kim
    • 1
  • Chang-Bum Lee
    • 1
  • Il-Kon Kim
    • 2
  1. 1.Dept. of Information and Network EngineeringYoungsan UniversityKorea
  2. 2.Dept. of Computer ScienceKyungpook National UniversityKorea

Personalised recommendations