An Improvement of Dissimilarity-Based Classifications Using SIFT Algorithm

  • Evensen E. Masaki
  • Sang-Woon Kim
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6744)


In dissimilarity-based classifications (DBCs), classifiers are not based on the feature measurements of individual objects, but rather on a suitable dissimilarity measure among the objects. In this paper, we study a new way of measuring the dissimilarity between two object images using a SIFT (Scale Invariant Feature Transformation) algorithm [5], which transforms image data into scale-invariant coordinates relative to local features based on the statistics of gray values in scale-space. With this method, we find an optimal or nearly optimal matching among differing images in scaling and rotation, which leads us to obtain dissimilarity representation after matching them. Our experimental results, obtained with well-known benchmark databases, demonstrate that the proposed mechanism works well and, compared with the previous approaches, achieves further improved results in terms of classification accuracy.


Face Recognition Object Image Dynamic Time Warping Scale Invariant Feature Transformation Benchmark Database 
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.
    Adini, Y., Moses, Y., Ullman, S.: Face recognition: the problem of compensating for changes in illumination direction. IEEE Trans. Pattern Anal. and Machine Intell. 19(7), 721–732 (1997)CrossRefGoogle Scholar
  2. 2.
    Duin, R.P.W., Juszczak, P., de Ridder, D., Paclik, P., Pekalska, E., Tax, D.M.J.: PRTools 4: a Matlab Toolbox for Pattern Recognition. Technical Report, Delft University of Technology, The Netherlands (2004),
  3. 3.
    Georghiades, A.S., Belhumeur, P.N., Kriegman, D.J.: From few to many: Illumination cone models for face recognition under variable lighting and pose. IEEE Trans. Pattern Anal. and Machine Intell. 23(6), 643–660 (2001)CrossRefGoogle Scholar
  4. 4.
    Kim, S.-W., Gao, J.: A dynamic programming technique for optimizing dissimilarity-based classifiers. In: da Vitoria Lobo, N., Kasparis, T., Roli, F., Kwok, J.T., Georgiopoulos, M., Anagnostopoulos, G.C., Loog, M. (eds.) S+SSPR 2008. LNCS, vol. 5342, pp. 654–663. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  5. 5.
    Lowe, D.G.: Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision 60(2), 91–110 (2004)CrossRefGoogle Scholar
  6. 6.
    Pekalska, E., Duin, R.P.W.: The Dissimilarity Representation for Pattern Recognition: Foundations and Applications. World Scientific Publishing, Singapore (2005)CrossRefzbMATHGoogle Scholar
  7. 7.
    Sim, T., Baker, S., Bsat, M.: The CMU pose, illumination, and expression (PIE) database of human faces. Technical report CMU-RI-TR-01-02, Robotics Institute, Carnegie Mellon University, Pittsburgh, PA (2001)Google Scholar
  8. 8.
    Sørensen, L., Loog, M., Lo, P., Ashraf, H., Dirksen, A., Duin, R.P.W., de Bruijne, M.: Image dissimilarity-based quantification of lung disease from CT. In: Jiang, T., Navab, N., Pluim, J.P.W., Viergever, M.A. (eds.) MICCAI 2010. LNCS, vol. 6361, pp. 37–44. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  9. 9.
    Wang, L., Zhang, Y., Feng, J.: On the Euclidean distance of images. IEEE Trans. Pattern Anal. and Machine Intell. 27(8), 1334–1339 (2005)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Evensen E. Masaki
    • 1
  • Sang-Woon Kim
    • 1
  1. 1.Dept. of Computer Science and EngineeringMyongji UniversityYonginSouth Korea

Personalised recommendations