Chapter 1: Invited Addresses and Tutorials on Signals, Coding,
  Systems and Intelligent Techniques

Title: An operator for the analysis of superimposed intrinsically two dimensional patterns

Author(s): Di Zang, Gerald Sommer

Address: Cognitive Systems Group, Institute of Computer Science and Applied Mathematics, Christian Albrechts University of Kiel, 24118 Kiel, Germany | Cognitive Systems Group, Institute of Computer Science and Applied Mathematics, Christian Albrechts University of Kiel, 24118 Kiel, Germany

Reference: 12th International Workshop on Systems, Signals and Image Processing pp. 77 - 81

Abstract/Summary: In this paper, we present an operator for the analysis of superimposed intrinsically two dimensional (i2D) patterns. Coupling the ideas of the monogenic signal, the tensor algebra and the quadrature filter, a rotationally invariant operator is derived. This operator is able to capture important features (e.g. amplitude, energy, orientation and phase) of the intrinsically 2D structure. Compared with other approaches, like the structure tensor and the boundary tensor, the main contribution of the proposed operator is the possibility to evaluate the phase information. The energy output is considered as a junction strength to detect the points of interest and the estimated orientation represents the main orientation of the local i2D structure. Experimental results illustrate the features extracted from the proposed operator. As an application, some test results of junction detection are also presented.

Order a copy of this article Order a copy of this article