• DocumentCode
    2409373
  • Title

    A recursive estimation of parameters of straight lines and circles: application to the segmentation of the Rey´s Complex Figure

  • Author

    Amara, Mounir ; deBrucq, D. ; Courtellemont, Pierre ; Wallon, Philippe ; Mesmin, C. ; Lecourtier, Yves

  • Author_Institution
    PSI-La3i, Rouen Univ., Mont-Saint-Aignan, France
  • Volume
    2
  • fYear
    1996
  • fDate
    25-29 Aug 1996
  • Firstpage
    467
  • Abstract
    We explain a new approach in on-line segmentation of handwritten drawings by estimating circle bows or linear segments. In these two cases, we use an extended Kalman filter. The state and observation equations come from cartesian equations of these geometrical shapes with no-varying parameters to estimate. The parameters are recursively estimated with a very good accuracy during the acquisition of the points. The method is tested on a particular kind of handwritten drawings: the Rey´s Complex Figure (Rey, 1942), used in neurological and psychological diagnosis in children. From the proposed segmentation, it should be easier to obtain the features retained by the psychologist in his analysis
  • Keywords
    Kalman filters; curve fitting; diagrams; filtering theory; image recognition; image segmentation; psychology; recursive estimation; Complex Figure; cartesian equations; children; circle bows; circles; curve fitting; extended Kalman filter; geometrical shapes; handwritten drawings; image segmentation; linear segments; neurological diagnosis; observation equations; online segmentation; psychological diagnosis; recursive parameter estimation; state equations; straight lines; Equations; Filtering; Kalman filters; Nonlinear filters; Parameter estimation; Psychology; Recursive estimation; Shape; State estimation; Testing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Pattern Recognition, 1996., Proceedings of the 13th International Conference on
  • Conference_Location
    Vienna
  • ISSN
    1051-4651
  • Print_ISBN
    0-8186-7282-X
  • Type

    conf

  • DOI
    10.1109/ICPR.1996.546869
  • Filename
    546869