Advertisement

Method of contour recognition

  • Z. M. Wójcik

Abstract

The method consists in an automatic conversion of an input image to a set of binary ones in such a manner that each binary image stands for objects of approximate brightness. Contours are detected as boundaries of all objects of each binary image and are represented by the digital value 1 in a memory.

A circular operator (field) is used to follow the contours detected. To recognize a contour segment inside the operator field, the central element and at least two nonadjacent peripheral elements of the operator must have the digital value of 1. The contour segment is represented by a very simple graph consisting of one node labeled with the word “segment” and several arcs joined to the node and labeled with the attributes of the feature “segment” such as “direction, a;” “coordinates, X, Y;” “length, r;” “straight” or “curved;” and “contour;” where a, X, Y, and r are measures of the attributes. The operator is then translocated along the contour and for each of its positions the same graph is obtained: direction and position measures may differ, two consecutive nodes being joined by means of an arc denoted with the name “to adjoin.” Reduction of every two consecutive nodes to a single node is carried out when their direction measures are the same.

The recognized contour (i.e., converted to a graph) is compared with a set of reference (standard) graphs to be identified. Standard (reference) graphs may be achieved as a result of recognition processes of reference patterns.

Keywords

Operator Field Gray Level Input Image Binary Image Central Element 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    M. Hueckel, “An operator which locates edges in digital pictures,” J.A.C.M., 15, 1972.Google Scholar
  2. 2.
    J. L. Muerle, D. C. Allen, “Experimental evaluation of techniques for automatic segmentation of objects in a complex scene,” in: Pictorial Pattern Recognition, G. C. Cheng (ed.), Thompson, Washington D.C., 1968.Google Scholar
  3. 3.
    M. Nowakowska, “Some psychological problems in psychometry, and quantitative approach to the dynamica of perception,” General Systems, XII, 1967.Google Scholar
  4. 4.
    A. Guzman, “Decomposition of a visual scene into three-dimensional bodies,” in: Automatic Interpretation and Classification of Images, A. Grasseli (ed.), Academic Press, 1969.Google Scholar
  5. 5.
    A. Rosenfeld, C. A. Kak, Digital Picture Processing, Academic Press, 1976.Google Scholar
  6. 6.
    J. Sklansky, “Image segmentation and feature extraction,” in: Digital Image Processing and Analysis, J. C. Simon and A. Rosenfeld (ed.) 1977, Nordhoff International Publishing, pp. 125–168.Google Scholar
  7. 7.
    Z. M. Wójcik, “A method for recognition of object position in robot system by means of contour analysis,” Silesia Technical University Reports, No. 56, Gliwice, 1980 (in Polish), pp. 185–201.Google Scholar
  8. 8.
    R. M. Haralick, “Edge and region analysis for digital image data,” C.G.I.P., 1, 1980.Google Scholar
  9. 9.
    A. Rosenfeld, M. Thurston, IEEE Trans. C-20, 1971, pp. 562–569.Google Scholar
  10. 10.
    Z. M. Wójcik, “A model of pattern recognition, and decidability of the natural language,” Proc. of the Intern. Symp. on System-Modelling-Control (organized by the Polish Cybernetic Society, Zakopane) 1979, Poland, pp. 383–401.Google Scholar
  11. 11.
    P. E. Danielsson, B. Kruse, “Distance checking algorithms,” Computer Graphics and Image Processing (C.G.I.P.), 4, 1979, pp. 349–376.CrossRefGoogle Scholar
  12. 12.
    J. S. M. Prewitt, and M. L. Mendelsohn, “The analysis of cell images,” Ann. N.Y. Acad. Sci., 128, 1966.Google Scholar
  13. 13.
    S. W. Zucker, “Algorithm for image segmentation,” in: Digital Processing and Analysis, J. C. Simon and A. Rosenfeld (ed.) 1977, Nordhoff Intern. Publishing, pp. 169–186.Google Scholar
  14. 14.
    H. Freeman, “On the encoding of arbitrary geometric configurations,” IEE Trans. Electronic Computers, Vol. EC-10, 2, 1961.Google Scholar
  15. 15.
    Z. M. Wójcik, “A system for an automatic detection of defects of semiconductor masks and printed circuit boards,” Electron Technology,4, 1977, Warsaw (published by The Institute of Electron Technology), pp. 95–108.Google Scholar
  16. 16.
    Z. M. Wójcik, “Automatic detection of semiconductor mask defects,” Microelectronics and Reliability, Vol. 15, 1976, pp. 585–593.CrossRefGoogle Scholar
  17. 17.
    L. E. Nordell, B. Kruse, “An adaptive operatorset,” Proc. of the 5th Int. Conf. on Pattern Recognition, Florida, 1980.Google Scholar
  18. 18.
    M. Nagao, T. Matsuyama, “Edge preserving smoothing,” Proc. of the Fourth Int. Conf. on Pattern Recognition, pp. 518–520, Kyoto, 1978, Japan.Google Scholar
  19. 19.
    Z. M. Wójcik, “A model of semantics of the natural language, and a fundamental condition for events representation processes and its applications,” Progress in Cybernetics and Systems Research,Vol. XI, R. Trappl et al. (ed.), Hemisphere Publishing Corporation (Washington, D.C.), pp. 257–268 and 381–392.Google Scholar

Copyright information

© Crane Russak & Company Inc 1984

Authors and Affiliations

  • Z. M. Wójcik
    • 1
  1. 1.The Industrial Electronics InstituteWarsawPoland

Personalised recommendations