Visual Form pp 39-46 | Cite as
Motion and Structure in Rigid Multi-Surfaced Stationary Environments Using Time-Varying Image Velocity: Linear Solutions
Abstract
This paper is concerned with the robust computation of motion and structure parameters, which describe an observer’s translation and rotation and the environmental structure, i.e. the depth of visible 3D points. Our new algorithms involve solving linear systems of equations relating these motion and structure parameters to time-varying image velocity. As in [9], our algorithms do not impose a local surface model but rather decouple depth scaled speed and the direction of translation from each other in the image velocity equation and then solve for each separately. However, since time-varying image velocity is used, an observer acceleration model is required.
Preview
Unable to display preview. Download preview PDF.
References
- [1]Adiv G., “Determining 3D Motion and Structure from Optical Flow Generated by Several Moving Objects”, COINS TR-84-07, University of Massachusetts, April, 1984.Google Scholar
- [2]Barron J., “Computing Motion and Structure From Noisy Time-Varying Image Velocity Information”, PhD Thesis, Dept. of Computer Science, University of Toronto, July, 1988 (also RBCV-TR-88-24, August, 1988).Google Scholar
- [3]Barron J.L., A.D. Jepson and J.K. Tsotsos, “The Feasibility of Motion and Structure from Noisy Time-Varying Image Velocity Information”, Vol. 5 No. 3, International Journal of Computer Vision, 1990, pp239-269.Google Scholar
- [4]Barron J.L., S. Beauchemin, T.A. Burkitt and D.J. Fleet, “A Comparison of Optical Flow Techniques”, submitted for publication, 1991.Google Scholar
- [5]Bruss A.R. and B.K.P. Horn, “Passive Navigation”, CVGIP (formally CGIP), Vol. 21, 1983, pp3–20.Google Scholar
- [6]Buxton B.F., H. Buxton, D.W. Murray and N.S. Williams, “3-D Solutions to the Aperture Problem” in Advances in Artifical Intelligence, T. O’Shea (editor), Elsevier Science Publishers B.V. (North Holland), 1984, pp631–640.Google Scholar
- [7]Fleet D.J. and A.D. Jepson, “Computation of Component Velocity from Local Phase Information”, International Journal of Computer Vision, 1990, pp77-104.Google Scholar
- [8]Heeger D.J., “Optical Flow Using Spatiotemporal Filters”, Intl. Journal of Computer Vision, Vol. 1, 1988, pp279–302.CrossRefGoogle Scholar
- [9]Heeger D.J. and A.D. Jepson, “A Simple Method for Computing 3D Motion and Depth”, ICCV, 1990, pp96-100 (see also “Subspace Methods for Recovering Rigid Motion 1: Algorithm and Implementation”, RBCV-TR-90-35, Dept. of Computer Science, University of Toronto, Nov., 1990).Google Scholar
- [10]Longuet-Higgins H.C. and K. Prazdny, “The Interpretation of a Moving Image”, Proc. of Royal Society of London, B208, 1980, pp385-397.Google Scholar
- [11]Lu X., “Motion and Structure Paradigms”, Masters Thesis (in preparation), Dept. of Computer Science, University of Western Ontario, 1991.Google Scholar
- [12]Prazdny K., “Motion and Structure From Optical Flow”, IJCAI 79, pp702-704.Google Scholar
- [13]Warren W. and D. Hannon, “Eye Movements and Optical Flow”, Journal of the Optical Society of America, Vol. 7, No. 1, 1990, pp 160–169.CrossRefGoogle Scholar
- [14]Waxman A.M., Kamgar-Parsi B. and M. Subbarao, “Closed-Form Solutions to Image Flow Equations for 3D Structure and Motion”, International Journal of Computer Vision, pp239-258, 1987.Google Scholar
- [15]Waxman A.M. and S. Ullman, “Surface Structure and Three-Dimensional Motion from Image Flow Kinematics”, Intl. Journal of Robotics Research, Vol. 4, No. 3, Fall, 1985, pp72–94.CrossRefGoogle Scholar
- [16]Waxman A.M. and K. Wohn, “Image Flow Theory: A Framework for 3D Inference from Time-Varying Imagery”, in Advances in Computer Vision, (editor C. Brown), Vol. 1, Lawrence Erlbaum Associates, 1988, pp165-224.Google Scholar