Unambiguous Determination of Velocity and Structure of an Accelerating Surface. A Theoretic Framework

  • Jens Arnspang
Conference paper
Part of the Informatik-Fachberichte book series (INFORMATIK, volume 181)


A theoretic framework for unambiguous determination of a surface, moving in the gravity of earth or another known acceleration field, is presented. Optic acceleration is defined and new motion constraint equations are derived. The scaling problem between surface depth and velocity is solved, also for non rigid surfaces. Determination of orientation and shape is discussed. All methods derived are local, linear and unambiguous.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    J. J. Gibson, The Perception of the Visual World, repr. by Greenwood Publishers 1974 (original 1950)Google Scholar
  2. [2]
    D. Marr, Vision, W. H. Freeman, San Francisco, 1982Google Scholar
  3. [3]
    J. J. Koenderink, Optic Flow, Vision Research, vol 26, no. 1, 1986Google Scholar
  4. [4]
    A. Waxman and K. Wohn, Image Flow Theory, Sensory Robotics Lab, Boston University, jan. 1986Google Scholar
  5. [5]
    B. K. P. Horn, Robot Vision, McGrawHill, 1986Google Scholar
  6. [6]
    B. K. P Horn and B. G. Schunck, Determination of Optical Flow, Artificial Intelligence, vol. 17, 1981Google Scholar
  7. [7]
    J. Arnspang, Notes on Local Determination of Optical Flow, DIKU Technical Report 88/1, 1988Google Scholar
  8. [8]
    K. Wohn et al, Motion Estimation based on multiple local Constraints, Pattern Recognition 16, 1983Google Scholar
  9. [9]
    A. Mitiche et al, Experiments in computing Optical Flow, Pattern Recognition, vol. 20, no. 2, 1987Google Scholar
  10. [10]
    H. H. Nagel, On the estimation on Optic flow, Artificial Intelligence, vol. 33, 1987Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1988

Authors and Affiliations

  • Jens Arnspang
    • 1
  1. 1.DIKU Computer Science DepartmentUniversity of CopenhagenCopenhagenDenmark

Personalised recommendations