A new possibility for pattern recognition independently from geometrical measuresetrical measures

  • Authors:
  • Z. Szakal;I. Zsoldos

  • Affiliations:
  • Faculty of Mechanical Engineering, Szent Istvan University, Godollo, Hungary;Faculty of Mechanical Engineering, Szent Istvan University, Godollo, Hungary

  • Venue:
  • ICOSSSE'08 Proceedings of the 7th WSEAS international conference on System science and simulation in engineering
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The base of this paper was a new method for the detection of exact and approximate reflective symmetry shown earlier [1]. The searching algorithm was based on the definition of the so called symmetry-parameter which is a rate of the symmetry, a number between 0 and 1 without a dimension and its value does not depend on geometrical measures. A so called shape-diagram was determined from the symmetry-parameters computed for various lines crossing the gravity centre. Beside the possibility of the symmetry detection, the shape-diagram shows an individual shape property of the 2D figures, independently from geometrical measures. In this paper we show a process in which similar 2D figures are sorted out from a multitude of different figures independently from geometrical measures using the definition of the shape-diagram.