Modelling Intermittently Present Features Using Nonlinear Point Distribution Models

  • Gerard Sanroma
  • Francesc Serratosa
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4872)


We present in this paper a new compact model to represent a data set based on the main idea of the Point Distribution Model (PDM). Our model overcomes PDM in two aspects, first, it is not needed all the objects to have the same number of points. This is a very important feature, since in real applications, not all the landmarks are represented in the images. And second, the model captures the nonlinearity of the data set. Some research have been presented that couples both aspects separately but no model have been presented until now that couples them as a whole. A case study shows the efficiency of our model and its improvement respect the models in the literature.


Point Distribution Models intermittently present land-marks missing data statistical shape modelling imputation 


  1. 1.
    Kass, M., Witkin, A., Terzopoulos, D.: Snakes - active contour models. International Journal of Computer Vision 1(4), 321–331 (1987)CrossRefGoogle Scholar
  2. 2.
    Cootes, T.F., Taylor, C.J., Cooper, D.H., Graham, J.: Active shape models - their training and application. Computer Vision and Image Understanding 61(1), 38–59 (1995)CrossRefGoogle Scholar
  3. 3.
    Sozou, P.D., Cootes, T.F., Taylor, C.J., Dimauro, E.C.: Nonlinear generalization of point distribution models using polynomial regression. Image and Vision Computing 13(5), 451–457 (1995)CrossRefGoogle Scholar
  4. 4.
    Rogers, M., Graham, J.: Structured point distribution models: Modelling intermittently present features. In: Proceedings of the British Machine Vision Conference, vol. 1, pp. 33–42 (2001)Google Scholar
  5. 5.
    Goodall, C.: Procrustes methods in the statistical-analysis of shape. Journal of the Royal Statistical Society Series B-Methodological 53(2), 285–339 (1991)zbMATHMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Gerard Sanroma
    • 1
  • Francesc Serratosa
    • 1
  1. 1.Dept. of Computer Engineering and Maths, Universitat Rovira i Virgili, Av. Països Catalans 26, E-43007 Tarragona, CataloniaSpain

Personalised recommendations