International Journal of Computer Vision

, Volume 22, Issue 1, pp 81–101 | Cite as

Gradient Based Image Motion Estimation Without Computing Gradients

  • Naresh Gupta
  • Laveen Kanal


Computing an optical flow field using the classical image motion constraint equation \(I_x u + Iy\upsilon + I_t = 0,\) is difficult owing to the aperture problem and the need to compute the image intensity derivatives via numerical differentiation—an extremely unstable operation. We integrate the above constraint equation over a significant spatio-temporal support and use Gauss's Divergence theorem to replace the volume integrals by surface integrals, thereby eliminating the intensity derivatives and numerical differentiation. We tackle the aperture problem by fitting an affine flow field model to a small space-time window. Using this affine model our new integral motion constraint approach leads to a robust and accurate algorithm to compute the optical flow field. Extensive experimentation confirms that the algorithm is indeed robust and accurate.

optical flow Gauss's Divergence theorem non-local constraint numerical differentiation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Anandan, P. 1989. A computational framework and an algorithm for the measurement of visual motion. International Journal of Computer Vision, 2:283–310.Google Scholar
  2. Barron, J., Fleet, D., and Beauchemin, S. 1994. Performance of optical flow techniques. International Journal of Computer Vision, 12(1):43–79.Google Scholar
  3. Burden, R., Faires, J., and Reynolds, A. 1978. Numerical Analysis. Prindle, Weber & Schmidt: Boston, MA.Google Scholar
  4. Fennema, C.L. and Thompson, W.B. 1979. Velocity determination in scenes containing several moving objects. Computer Graphics and Image Processing, 9:301–315.Google Scholar
  5. Fleet, D.J. and Jepson, A.D. 1990. Computation of component image velocity from local phase information. International Journal of Computer Vision, 5:77–104.Google Scholar
  6. Glazer, F. 1987. Hierarchical motion detection, Ph.D. Dissertation, University of Massachusetts, Amherest, MA.Google Scholar
  7. Gupta, N. 1993. Recovering Shape and Motion from a Sequence of Images, Ph.D. Dissertation, University of Maryland, College Park, Md, 1993.Google Scholar
  8. Gupta, N., Raghavan, S., and Kanal, L. 1993. Robust recovery of image motion. In Proc. Asian Conference on Computer Vision, Osaka, Japan.Google Scholar
  9. Gupta, N. and Kanal, L. 1995. 3D motion estimation from motion field, Artificial Intelligence, 78:45–86.Google Scholar
  10. Hay, J.C. 1960. Optical motions and space perception: An extension of Gibson's analysis. Psychological Review, 73(6):550– 565.Google Scholar
  11. Heeger, D.J. 1988. Optical flow using spatiotemporal filters. International Journal of Computer Vision, 1:273–302.Google Scholar
  12. Hildreth, E.C. 1984. Computations underlying the measurement of visual motion. Artificial Intelligence, 23:309–354.Google Scholar
  13. Horn, B.K.P. and Schunck, B.G. 1981. Determining optical flow. Artificial Intelligence, 17:185–203.Google Scholar
  14. Horn, B.K.P. 1986. Robot Vision. MIT Press.Google Scholar
  15. Kreyszig, E. 1967. Advanced Engineering Mathematics, John Wiley & Sons, Inc.Google Scholar
  16. Longuet-Higgins H.C. and Prazdny, K. 1980. The interpretation of a moving retinal image. Proc. Royal Soc. of London, B208:385–397.Google Scholar
  17. Longuet-Higgins, H.C. 1981. A computer algorithm for reconstructing a scene from two projections. Nature, 293:133–135.Google Scholar
  18. Lucas, B.D. 1984. Generalized image matching by the method of differences, Ph.D. Dissertation, Dept. of Computer Science, Carnegie-Mellon University.Google Scholar
  19. Nagel, N.N. 1981. Representation of moving rigid objects based on visual observations. IEEE Computer, pp. 29–39.Google Scholar
  20. Nagel, H.H. and Enkelmann, W. 1986. An investigation of smoothness constraints for the estimation of displacement vector fields from image sequences. IEEE Trans. PAMI, 8:565–593.Google Scholar
  21. Negahdaripour, S. and Lee, S. 1992. Motion recovery from image sequences unsing only first order optical flow information. International Journal of Computer Vision, 9(3):163–184.Google Scholar
  22. Poggio, T. and Reichardt, W. 1976. Visual control of orientation behavior in the fly: Part II towards underlying neural interactions, Quarterly Reviews in Biophysics, 9:377–438.Google Scholar
  23. Raghavan, S., Gupta, N., and Kanal, L. 1992. Discontinuity preserved image motion estimation. In Proc. International Conference on Pattern Recognition, The Hague, The Netherlands.Google Scholar
  24. Schunck, B.G. 1984. Motion segmentation by constraint line clustering. In Proc. IEEE Workshop on Computer Vision: Representation and Control, pp. 58–62.Google Scholar
  25. Tsai, R.Y. and Huang, T.S. 1981. Estimating 3-D motion parameters of a rigid planar patch. IEEE Trans. ASSP, 29(6):1147– 1152.Google Scholar
  26. Verri, A. and Poggio, T. 1987. Against quantitative optical flow. In Proc. International Conference on Computer Vision, pp. 171–180.Google Scholar
  27. Waxman, A.M. and Ullman, S. 1985. Surface structure and three-dimensional motion from image flow kinematics. Intern. J. Robot. Res., 4:95–108.Google Scholar

Copyright information

© Kluwer Academic Publishers 1997

Authors and Affiliations

  • Naresh Gupta
    • 1
    • 2
  • Laveen Kanal
    • 1
  1. 1.Department of Computer ScienceUniversity of Maryland
  2. 2.L.N.K. Corporation, Inc.Riverdale

Personalised recommendations