Multiple Classifier System Approach to Model Pruning in Object Recognition

  • Josef Kittler
  • Ali R. Ahmadyfard
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3024)


We propose a multiple classifier system approach to object recognition in computer vision. The aim of the approach is to use multiple experts successively to prune the list of candidate hypotheses that have to be considered for object interpretation. The experts are organised in a serial architecture, with the later stages of the system dealing with a monotonically decreasing number of models. We develop a theoretical model which underpins this approach to object recognition and show how it relates to various heuristic design strategies advocated in the literature. The merits of the advocated approach are then demonstrated experimentally using the SOIL database. We show how the overall performance of a two stage object recognition system, designed using the proposed methodology, improves. The improvement is achieved in spite of using a weak recogniser for the first (pruning) stage. The effects of different pruning strategies are demonstrated.


Object Recognition Recognition Rate Pruning Strategy Multiple Expert Candidate Label 
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.
    Ahmadyfard, A., Kittler, J.: Enhancement of ARG object recognition method. In: Proceeding of 11 the European Signal Processing Conference, September 2002, vol. 3, pp. 551–554 (2002)Google Scholar
  2. 2.
    Belongie, S., Malik, J., Puzicha, J.: Shape matching and object recognition using shape contexts. IEEE Transactions on Pattern Analysis and Machine Intelligence, 509–522 (2002)Google Scholar
  3. 3.
    Christmas, W.J., Kittler, J., Petrou, M.: Structural matching in computer vision using probabilistic relaxation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 749–764 (1995)Google Scholar
  4. 4.
    Fairhurst, M.C., Rahman, A.F.R.: Generalised approach to the recognition of structurally similar handwritten characters using multiple expert classifiers. IEE Proceeding on Vision, Image and Signal Processing 144(1), 15–22 (1997)CrossRefGoogle Scholar
  5. 5.
    Finlayson, G., Funt, B., Barnard, J.: Color constant color indexing. IEEE Transactions on Pattern Analysis and Machine Intelligence 17(5), 522–529 (1995)CrossRefGoogle Scholar
  6. 6.
  7. 7.
    Ianakiev, K., Govindaraju, V.: Architecture for classifier combination using entropy measures. In: Kittler, J., Roli, F. (eds.) MCS 2000. LNCS, vol. 1857, pp. 340–350. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  8. 8.
    Kim, G., Govindaraju, V.: A lexicon driven approach to handwritten word recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence 19(4), 366–379 (1997)CrossRefGoogle Scholar
  9. 9.
    Krassimir, I., Govindaraju, V.: An architecture for classifier combination using entropy measure. In: Kittler, J., Roli, F. (eds.) MCS 2000. LNCS, vol. 1857, pp. 340–350. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  10. 10.
    Lowe, D.G.: Three-dimensional object recognition form single two-dimensional image. Artificial Intelligence, 355–395 (1987)Google Scholar
  11. 11.
    Madhvanath, S., Kleinberg, E., Govindaraju, V.: Holistic verification for handwritten phrases. IEEE Transactions on Pattern Analysis and Machine Intelligence 21(12), 1344–1356 (1999)CrossRefGoogle Scholar
  12. 12.
    Matas, J(G.), Koubaroulis, D., Kittler, J.: Colour image retrieval and object recognition using the multimodal neighbourhood signature. In: Vernon, D. (ed.) ECCV 2000. LNCS, vol. 1842, pp. 48–64. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  13. 13.
    Murase, H., Nayar, S.: Visual learning and recognition of 3d objects from appearance. International Journal of Computer Vision, 5–24 (1995)Google Scholar
  14. 14.
    Rahman, A.F.R., Fairhurst, M.C.: Exploiting second order information to design a novel multiple expert decision combination platform for pattern classification. Electronic Letters 33, 476–477 (1997)CrossRefGoogle Scholar
  15. 15.
    Rahman, A.F.R., Fairhurst, M.C.: A new hybrid approach in combining multiple experts to recognise handwritten numerals. Pattern Recognition Letters 18, 781–790 (1997)CrossRefGoogle Scholar
  16. 16.
    Rahman, A.F.R., Fairhurst, M.C.: An evaluation of multi-expert configurations for for the recognition of handwritten numerals. Pattern Recognition 31, 1255–1273 (1998)CrossRefGoogle Scholar
  17. 17.
    Rahman, A.F.R., Fairhurst, M.C.: Enhancing multiple expert decision combination strategies through exploitation of a priori information sources. IEE Proceeding on Vision Image and Signal Processing 146, 40–49 (1999)CrossRefGoogle Scholar
  18. 18.
    Schmid, C., Mohr, R.: Local grayvalue invariants for image retrieval. IEEE Transactions on Pattern Analysis and Machine Intelligence 19(5), 530–535 (1997)CrossRefGoogle Scholar
  19. 19.
    Shao, Z., Kittler, J.: Shape representation and recognition using invariant unary and binary relations. Image and Vision Computing 17, 429–444 (1999)CrossRefGoogle Scholar
  20. 20.
    Swain, M.J., Ballard, D.H.: Colour indexing. Intl. Journal of Computer Vision 7(1), 11–32 (1991)CrossRefGoogle Scholar
  21. 21.
    Turner, M., Austin, J.: A neural relaxation technique for chemical graph matching. In: Proceeding of Fifth International Conference on Artificial Neural Networks (1997)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Josef Kittler
    • 1
  • Ali R. Ahmadyfard
    • 1
  1. 1.Centre for Vision, Speech and Signal ProcessingUniversity of SurreyGuildfordUnited Kingdom

Personalised recommendations