Advertisement

Adaptive hierarchical indexing and constrained localization: Matching characteristic views

  • Gunter Bellaire
  • Mathias Lübbe
Session IA2a — 3-D Image Analysis
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1024)

Abstract

This article presents a complete hybrid object recognition system for thredimensional objects using the characteristic view (ChV) idea. To apply the ChV representation method in a recognition system investigations are needed concerning the processing of large object data bases. First we present two methods to reduce the number of views in the object data base. Second we developed an accumulator (AC)-based matching strategy combined with a localization process. This strategy bases on a hierarchical indexing structure that uses a Gaussian distributed voting. The off-line part of the matching includes a statistical analysis of the object data base and an interface to process results of a sensor configuration analysis. The calculated results support the construction of an adapted layer model suitable for hierarchical indexing. Further an unsupervised learning module is introduced, that investigates the measurement errors and adapts the system online. Results of the matching are verified by a localization tool, which uses an interpretation tree search combined by a shape from angle method and a constrained alignment technique. The article shows results with real greyscale images.

Keywords

Photometric Stereo Sensor Configuration Scene Object Characteristic View Junction Type 
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.
    F. Arman, J. K. Aggarwal, “CAD-Based Vision: Object Recognition in Cluttered Range Images Using Recogniton Strategies”, CVGIP: Image Understanding, Vol. 58, no. 1, July, pp. 33–48, 1993Google Scholar
  2. 2.
    G. Bellaire, “Feature-Based Computation of Hierarchical Aspect-Graphs”, Machine GRAPHICS & VISION, Vol. 2, no 2, pp. 105–122, 1993Google Scholar
  3. 3.
    G. Bellaire, K. Schlüns, A. Mitritz, K. Gwinner, “Adaptive Matching Using Object Models Generated From Photometric Stereo Images”, to appear in Proc. of 8th ICIAP, San Remo, Italy, 1995Google Scholar
  4. 4.
    G. Bellaire, “Hashing with a Topological Invariant Feature”, to appear in Proc. of ACCV'95, SingaporeGoogle Scholar
  5. 5.
    K.W. Bowyer (Org) “Why Aspect Graphs are not (yet) Practikal for Computer Vision”, Workshop Panel Report, CVGIP, Vol. 55, no. 2, March, pp. 97–105, 1992Google Scholar
  6. 6.
    I. Chakravarty, H. Freeman, “Characteristic Views as a Basis for 3-D Object Recognition”, Proc. Soc. Photo-Optical Instrumentation, Robot Vision, Vol. 336, SPIE Bellingham, WA, pp. 37–45, 1982Google Scholar
  7. 7.
    S. J. Dickinson, A. Pentland, A. Rosenfeld, “3D Shape Recovery Using Distributed Aspect Matching”, IEEE Trans. on PAMI, Vol. 14, no. 2, pp. 174–197, 1992Google Scholar
  8. 8.
    D. E. Eggert, K. W. Bowyer, Ch. R. Dyer, H. I. Christensen, D. B. Goldgof, “The Scale Space Aspect Graph”, Proc. CVPR, pp. 335–340, 1992Google Scholar
  9. 9.
    P. Flynn, A.K. Jain, “CAD-Based Computer Vision: From CAD Models to Relational Graphs”, IEEE Trans. on PAMI, Vol. 13, no. 2, Feb, pp. 114–132, 1991Google Scholar
  10. 10.
    Z. Gigus, J. Malik, “Computing the aspect for line drawings of polyhedral objects, “IEEE Trans. on PAMI, Vol. 12 no. 2, pp. 113–123, 1990Google Scholar
  11. 11.
    D. P. Huttenlocher, S. Ullman, “Recognizing Solid Objects by Alignment with an Image”, IJCV, Vol. 5, no. 2, pp. 195–212, 1990Google Scholar
  12. 12.
    K. Ikeuchi, “Determining Linear Shape Change: Toward Automatic Generation of Object Recognition Programs”, CVGIP, Vol. 53, pp. 154–170, 1991Google Scholar
  13. 13.
    K. Kanatani, “Group-theoretical Methods in Image Understanding”, Springer, Heidelberg, 1990Google Scholar
  14. 14.
    H. Lu, L. G. Shapiro, “Model-Based Vision Using Relational Summaries”, SPIE Vol. 1095, Applications of Artificial Intellience VII, pp. 662–675, 1989Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  • Gunter Bellaire
    • 1
  • Mathias Lübbe
    • 1
  1. 1.Computer Vision Group, Dep. of Computer ScienceTechnical University of BerlinBerlinGermany

Personalised recommendations