Two-dimensional object recognition using a two-dimensional polar transform

https://doi.org/10.1016/0031-3203(91)90007-RGet rights and content

Abstract

A two-dimensional (2D) transform is proposed for the classification of planar objects. With a centroid referenced polar representation it samples the multiple intersections of N radii with the object. Using the mass center, translation invariance is achieved and, with an appropriate amplitude normalization, the transform is also made invariant to scaling. The modules of its 2D Fast Fourier Transform coefficients, which are rotation invariant, are used as the feature vector. Simultaneously, translation, scaling and rotation parameters are estimated. The high performance of this algorithm is shown with computer simulations, and its advantages and disadvantages are analysed in comparison with existing methods.

References (32)

  • G.C. Stockman et al.

    2D object acquisition using circular scanning

    Pattern Recognition

    (1984)
  • C.T. Zahn et al.

    Fourier descriptors for plane closed curves

    IEEE Trans. Comput.

    (1972)
  • G.H. Granlund

    Fourier preprocessing for hand print character recognition

    IEEE Trans. Comput.

    (1972)
  • E. Persoon et al.

    Shape discrimination using Fourier descriptors

    IEEE Trans. Syst. Man Cybern.

    (1977)
  • S. Maitra

    Moment invariants

  • Y.S. Abu-Mostafa et al.

    Image normalization by complex moments

    IEEE Trans. Pattern Anal. Mach. Intell.

    (1985)
  • Cited by (26)

    • The spirals of the Slope Chain Code

      2019, Pattern Recognition
      Citation Excerpt :

      The chain elements produce finite alphabets which allow us to use grammatical techniques for shape classification. Most chain code methods presented in literature [2,10–14] are based on the representation of contour shapes by means of constant straight-line segments at different previous-defined directions. In 2D domain, there are methods to represent 2D curves via chain coding for example in refs. [2,10,12–15].

    • A chain code for representing high definition contour shapes

      2019, Journal of Visual Communication and Image Representation
      Citation Excerpt :

      Thus, chain code techniques may be a useful tool for representing contour shapes in computer vision and pattern recognition. Most chain code methods presented in literature [1–6] are based on the representation of contour shapes by means of constant straight-line segments at different previous-defined directions. The above-mentioned methods produce low definition contour shapes.

    • A measure of tortuosity based on chain coding

      2013, Pattern Recognition
      Citation Excerpt :

      The result is a chain for every curve, where each element of the chain represents the slope change at a given point. The SCC is similar to other chain codes [7,12,13] since it uses numerical sequences, but has some important differences from them. The main characteristics of the SCC are: independence of translation, rotation, and optionally, of scaling; it does not use a grid; the straight-line segment size (l) is always the same for the whole shape; and the range of slope changes is unlimited (goes continuously from −1 to 1).

    • Efficiency of chain codes to represent binary objects

      2007, Pattern Recognition
      Citation Excerpt :

      Classical methods for processing chains are referred to [4]. Other interesting coding schemes that are related to chain code are available in Refs. [5–9]. The vertex chain code (VCC) was presented in 1999 [10].

    • Shape matching and recognition using a physically based object model

      2001, Computers and Graphics (Pergamon)
      Citation Excerpt :

      In [13–15], the boundary of a region is represented by a sequence of numbers and the shape matching is accomplished by string matching. Other shape-matching methods involve finding the polar transform of the shape sample [16] or calculating the distances of the feature points from the centroid [17], etc. In [18], the shape matching is accomplished by graph matching for multilevel structural descriptions of shape samples.

    View all citing articles on Scopus
    View full text