Recognizing partially visible 2-D non-rigid wire-shapes

  • J. H. Sossa-Azuela
  • A. E. de León-Gutierrez
Poster Papers
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1451)


A technique for the fast recognition of partially visible 2-D non-rigid wire-shapes is presented. For this, the skeleton of an object is first decomposed into a set of parts called fundamental parts (FPs). Each FP is next described by a code composed of five integer numbers. The code of each FP is then used to build the corresponding database of models. During a second phase (of indexing), the FPs extracted from a test image are used to index the model database subtracting from it the list of objects possibly present in the image. The performance of the system is tested with a set of insect designs.


  1. [1]
    J. H. Sossa and J. Díaz-de-León S., Recognizing 2-D rigid and non-rigid wire-shapes, in Advances in Structural and Syntactical Pattern Recognition, Lecture Notes in Computer Science, 1121. Springer, 300–309, 1996.Google Scholar
  2. [2]
    R. Horaud and J. H. Sossa, Polyhedral object recognition by indexing, Pattern Recognition, 28(12), 1995, pp. 1855–1670.CrossRefGoogle Scholar
  3. [3]
    J. Díaz-de-León S., Skeletonizing algorithms for binary images, Master's Thesis, CINVESTAV-IPN. Electrical Engineering Department, Mexico, D.F., August de 1993. In Spanish.Google Scholar
  4. [4]
    J. L. Díaz de León S. and J. H. Sossa, On the computation of the Euler number of a binary object. Pattern Recognition, 29(3), 1996, pp. 471–476.CrossRefGoogle Scholar
  5. [5]
    R. Merris, The second immanantal polynomial and the centroid of a graph, SIAM J. Alg. Disc. Meth. 7, 1986, pp. 484–503.Google Scholar
  6. [6]
    G. M. Constantine, Graph complexity and the laplacian matrix in blocked environments, Linear and Multilinear Alg. 28, 1990, pp. 49–56.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • J. H. Sossa-Azuela
    • 1
  • A. E. de León-Gutierrez
    • 1
  1. 1.Centre de Investigación en Computación-IPNUnidad Profesional Adolfo López MateosMéxicoMéxico

Personalised recommendations