New Variants of the SDF Classifier

  • Maciej Smiatacz
  • Witold Malina
Part of the Advances in Intelligent and Soft Computing book series (AINSC, volume 57)


This paper is addressing problems related to the construction of classifiers in which the typical vector representation of a pattern is replaced with matrix data. We introduce some new variants of the method based on the Similarity Discriminant Function (SDF). The algorithms were tested on images of handwritten digits and on the photographs of human faces. On the basis of the experiments we can state that our modifications improved the performance of the SDF classifier.


Feature Vector Recognition Rate Handwritten Digit Face Recognition System Projection Axis 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Turk, M., Pentland, A.: Eigenfaces for Recognition. J. Cognitive Neuroscience 3(1), 71–86 (1990)CrossRefGoogle Scholar
  2. 2.
    Cheng, Y.-Q., Liu, K., Yang, J.-Y.: A Novel Feature Extraction Method for Image Recognition Based on Similar Discriminant Function (SDF). Pattern Recognition 26(1), 115–125 (1993)CrossRefGoogle Scholar
  3. 3.
    Cheng, Y.-Q., Zhuang, Y.-M., Yang, J.-Y.: Optimal Fisher Discriminant Analysis Using the Rank Decomposition. Pattern Recognition 25(1), 101–111 (1992)CrossRefMathSciNetGoogle Scholar
  4. 4.
    Samaria, F., Harter, A.: Parameterisation of a Stochastic Model for Human Face Identification. In: Proc. 2nd IEEE Workshop on Applications of Computer Vision (1994)Google Scholar
  5. 5.
    Garris, M.D., Wilkinson, R.A.: NIST Special Database 3. Handwritten Segmented Characters. National Institute of Standard and Technology, Gaithesburg, MD, USA (1992)Google Scholar
  6. 6.
    Okada, T., Tomita, S.: An Optimal Orthonormal System for Discriminant Analysis. Pattern Recognition 18(2), 139–144 (1985)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Maciej Smiatacz
    • 1
  • Witold Malina
    • 1
  1. 1.Faculty of Electronics, Telecommunications and InformaticsGdansk University of TechnologyGdanskPoland

Personalised recommendations