Advertisement

Measurement and integration of 3-D structures by tracking edge lines

  • James L. Crowley
  • Patrick Stelmaszyk
Tracking
Part of the Lecture Notes in Computer Science book series (LNCS, volume 427)

Abstract

This paper describes a technique for building a geometric description of a scene from the motion of a camera mounted on a robot arm. The movements of edge-lines in a sequence of image are tracked to maintain an image plane "flow model". Tracking perserves the correspondance of segments, even when the camera displaces, makeing possible a inexpensive form of motion stereo. Three dimensional structure is computed using the matches provided by the segment tracking process and the displacement parameters provided by the robot controller. By fusion of 3D data from different view points, we obtain an accurate and complete representation of the scene.

Results from a sequence of 80 images taken from a camera mounted on a robot arm are presented to illustrate the technique. These results are used for an experimental evaluation ito illustrate the accuracy and the robustness of the technique.

Keywords

Kalman Filter Flow Model Composite Model Model Segment Tracking Process 
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.

References

  1. [Ayache 87]
    Ayache, A. Faugeras, O. Maintaining Representation of the Environment of a Mobile Robot. In proc. International Symposium on Robotics Research, Santa Cruz, California, USA, August 1987.Google Scholar
  2. [Ayache 89]
    Ayache, N. "Construction et Fusion de Représentations Visuelles 3D", Thèse de Doctorat d'Etat, Universtié Paris-Sud, centre d'Orsay, 1988Google Scholar
  3. [Brammer 89]
    Brammer K. and G. Siffling, Kalman Bucy Filters, Artech House Inc., Norwood MA, USA, 1989.Google Scholar
  4. [Canny 86]
    Canny, J. "A Computational Approach to Edge Detection", IEEE Trans. on P.A.M.I., Vol PAMI-8, No. 6, Nov. 1986.Google Scholar
  5. [Chehikian 88]
    Stelmaszyk, P., C. Discours, and, A. Chehikian, "A Fast and Reliable Token Tracker", In IAPR Workshop on Computer Vision, Tokyo, Japan, October, 1988.Google Scholar
  6. [Chehikian 89]
    Chehikian, A., S. Stelmaszyk and P. Depaoli, "Hardware Evaluation Process for tracking edges lines.", Workshop on Industrial Applications of Machine Intel. and Vision. Tokyo, 1989.Google Scholar
  7. [Crowley 85]
    Crowley, J. L., "Navigation for an Intelligent Mobile Robot", IEEE Journal on Robotics and Automation, 1 (1), March 1985.Google Scholar
  8. [Crowley 86]
    Crowley, J. L., "Representation and Maintenance of a Composite Surface Model", IEEE International Conference on Robotics and Automation, San Francisco, Cal., April, 1986.Google Scholar
  9. [Crowley-Ramparany 87]
    Crowley, J. L. and F. Ramparany, Mathematical Tools for Manipulating Uncertainty in Perception", AAAI Workshop on Spatial Reasoning and Multi-Sensor Fusion", Kaufmann Press, October, 1987.Google Scholar
  10. Crowley et. al. 88]
    Crowley, J. L., P. Stelmaszyk and C. Discours, "Measuring Image Flow by Tracking Edge-Lines", ICCV 88: 2nd International Conference on Computer Vision, Tarpon Springs, Fla., Dec. 1988.Google Scholar
  11. [Deriche 87]
    Deriche, R., "Using Canny's Criteria to Derive a Recursively Implemented Optimal Edge Detector", International Journal of Computer Vision, Vol 1(2), 1987.Google Scholar
  12. [Durrant-Whyte 87]
    Durrant-Whyte, H. F., "Consistent Integration and Propagation of Disparate Sensor Observations", Int. Journal of Robotics Research, Spring, 1987.Google Scholar
  13. [Faugeras, et. al. 86]
    Faugeras, O. D., N. Ayache, and B. Faverjon, "Building Visual Maps by Combining Noisey Stereo Measurements", IEEE International Conference on Robotics and Automation, San Francisco, Cal., April, 1986.Google Scholar
  14. [Faugeras and Toscani]
    Faugeras, O. D., G. Toscani, “The Calibration Problem for Stereo. Computer Vision and Pattern Recognition, pp 15–20, Miami Beach, Florida, USA, June 1986.Google Scholar
  15. [Gennery, 82]
    Gennery, D. B., "Tracking Known Three Dimensional Objects", Proc. of the National Conference on Artificial Intelligence (AAAI-82), Pittsburgh, 1982.Google Scholar
  16. [Harris 1988]
    Harris C.G, "Using a Sequence of More than 2 Images", IEEE Workshop on Motion and Stereopsis in Machine Vision", Digest 7, pp 44–49, 1989.Google Scholar
  17. [Huang 83]
    Huang T. H., Image Sequence Processing and Dynamic Scene Analysis, Springer Verlag, Berlin, 1983.Google Scholar
  18. [Hildreth 82]
    Hildreth, E. C., The Measurement of Visual Motion, 1983 ACM Distinguished Dissertation, MIT Press, Cambridge Mass. 1982.Google Scholar
  19. [Matthies et. al. 87]
    Matthies, L., R. Szeliski, and T. Kanade, "Kalman Filter-based Algorithms for Estimating Depth from Image Sequences", CMU Tech. Report, CMU-CS-87-185, December 1987.Google Scholar
  20. [Ramaparany 89]
    Ramparany, F., "Perception Multi-sensorielle de la Structure Geometrique d'une Scene", Thèse de Doctrat, INPG, Feb 1989.Google Scholar
  21. [Roach and Aggarwal 80]
    Roach, J. W. and J. K. Aggarwal, "Determining the Movement of Objects in a Sequence of Images", IEEE Transactions on P.A.M.I., PAMI-2, No. 2, 1980.Google Scholar
  22. [Shui-Ahmad 87]
    Shui, Y. C. and S. Ahmad, "Finding the mounting position of a sensor by solving a homogeneous transformation equation of the form AX = XB", in Proc. of the IEEE International Conference of Robotics and Automation, San Francisco, June, 1986.Google Scholar
  23. [Suzuki and Yachida 89]
    Suzuki k. and Yachida M, "Establishing Correspondence and Getting 3-D Information in Dynamic Images", On Electronic Information and Communication Journal, Vol J72-D-II No5, pp 686–695 May 1989 (In Japanese).Google Scholar
  24. [Tsai 85]
    Tsai, R. Y., "An Efficient and Accurate Camera Calibration Technique for 3-D Machine Vision", Technical Report, IBM T. J. Watson Reearch Center, 1986.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1990

Authors and Affiliations

  • James L. Crowley
    • 1
  • Patrick Stelmaszyk
    • 2
  1. 1.LIFIA (CNRS) - IMAGUSA
  2. 2.I.T.M.I.USA

Personalised recommendations