Regular Article
Robust and Efficient Fourier–Mellin Transform Approximations for Gray-Level Image Reconstruction and Complete Invariant Description

https://doi.org/10.1006/cviu.2001.0922Get rights and content

Abstract

This paper addresses the gray-level image representation ability of the Fourier–Mellin transform (FMT) for pattern recognition, reconstruction, and image database retrieval. The main practical difficulty of the FMT lies in the accuracy and efficiency of its numerical approximation and we propose three estimations of its analytical extension. A comparison of these approximations is performed from discrete and finite-extent sets of Fourier–Mellin harmonics by means of experiments in: (i) image reconstruction via both visual inspection and the computation of a reconstruction error; and (ii) pattern recognition and discrimination by using a complete and convergent set of features invariant under planar similarities.

Experimental results on real gray-level images show that it is possible to recover an image to within a specified degree of accuracy and to classify objects reliably even when a large set of descriptors is used. Finally, an example will be given, which illustrates both theoretical and numerical results in the context of content-based image retrieval.

References (37)

  • M. Daoudi et al.

    Shape distances for contour tracking and motion estimation

    Pattern Recognition

    (1999)
  • H. Ming-Kuel

    Visual pattern recognition by moment invariants

    IRE Trans. Inform. Theory

    (1962)
  • S.S. Reddi

    Radial and angular moment invariants for image identification

    IEEE Trans. Pattern Anal. Mach. Intell.

    (1981)
  • Y. Abu-Mostafa et al.

    Image normalisation by complex moments

    IEEE Trans. Pattern Anal. Mach. Intell.

    (1985)
  • I. Rothe et al.

    The method of normalisation to determine invariants

    IEEE Trans. Pattern Anal. Mach. Intell.

    (1996)
  • M. Pawlak

    On the reconstruction aspects of moment descriptors

    IRE Trans. Inform. Theory

    (1992)
  • M.R. Teague

    Image analysis via the general theory of moments

    J. Opt. Soc. Amer. A

    (1980)
  • C.H. Teh et al.

    On image analysis by the method of moments

    IEEE Trans. Pattern. Anal. Mach. Intell.

    (1988)
  • Cited by (144)

    • Implementing the homography-based global HR-EBSD/TKD approach

      2022, Advances in Imaging and Electron Physics
      Citation Excerpt :

      Therefore, the latter algorithm is applied first. Ways to approximate the FMT-CC are reviewed in (Derrode & Ghorbel, 2001). At this stage of development, the method by (Reddy & Chatterji, 1996) is chosen because of its simplicity.

    • Global DIC approach guided by a cross-correlation based initial guess for HR-EBSD and on-axis HR-TKD

      2020, Acta Materialia
      Citation Excerpt :

      Robustness against noise or pattern blurring can be enhanced by means of error functions [66]. As far as the initial guess is concerned, the FMT is currently approximated but an analytical computation is also possible as reviewed [67]. More generally, the initial guess strategy can be optimised by coupling the present path independent method with a reliability guided and path dependant one.

    • An almost complete curvature scale space representation: Euclidean case

      2019, Signal Processing: Image Communication
      Citation Excerpt :

      The multi-scale Fourier-based descriptor proposed by [28] represented the shape using its boundary and its content using the Gaussian filter in many scales. It is E(2) invariant and robust to noise. [29] introduced the analytical Fourier–Mellin transform which gave also an invariant description for the gray-level image.

    • Optimizing the Extended Fourier Mellin Transformation Algorithm

      2023, IEEE International Conference on Intelligent Robots and Systems
    View all citing articles on Scopus
    View full text