Recursive affine structure and motion from image sequences
This paper presents a new algorithm for structure from motion from an arbitrary number of tracked features over an arbitrary number of images, which possesses several advantages over previous formulations. First, it is recursive, so the time complexity is independent of the number of images. The complexity is linear with the number of tracked features. The algorithm allows newly appeared features to be included, stale features to be discarded, and missing data to be handled naturally. Dynamic outlier elimination is achieved without recourse to heuristic segmentation strategies. Lastly, the algorithm can employ different kinds of tracked features, e.g. edges and corners, in the same framework.
The actual structure from motion recovered is affine, which assumes limited depth variation within the field of view, but the recovery is based on a more general recursive estimation algorithm, known as the variable state dimension filter (VSDF), which we devised and applied earlier to active camera calibration.
Results are presented for real image sequences, and timings for the algorithm demonstrate the feasibility for real-time implementation.
KeywordsStructure from motion affine invariance recursive filter
- 1.K. J. Bradshaw, P. F. McLauchlan, I. D. Reid, and D. W. Murray. Saccade and pursuit on an active head/eye platform. In J. Illingworth, editor, Proc. 4th British Machine Vision Conf., Guildford. BMVA Press, 1993. to appear in Image and Vision Computing Special Issue on BMVC '93.Google Scholar
- 2.S. Demey, A. Zisserman, and P. Beardsley. Affine and projective structure from motion. In D. Hogg and R. Boyle, editors, Proc. 3rd British Machine Vision Conf., Leeds, pages 49–58. Springer-Verlag, September 1992.Google Scholar
- 3.C. J. Harris and M. Stephens. A combined corner and edge detector. In Proc. 4th Alvey Vision Conf., Manchester, pages 147–151, 1988.Google Scholar
- 4.J. J. Koenderink and A. J. van Doorn. Affine structure from motion. J. Opt. Soc. Am. A, 8(2):377–385, 1991.Google Scholar
- 5.P. F. McLauchlan. Horatio: libraries for vision applications. Technical Report OUEL 1967/92, Dept. Engineering Science, University of Oxford, October 1992.Google Scholar
- 6.P. F. McLauchlan and D. W. Murray. Variable state dimension filter applied to active camera calibration. In Proc SPIE Sensor Fusion VI, Boston MA, pages 14–25, September 1993.Google Scholar
- 7.P.F. McLauchlan and D.W. Murray. Active camera calibration for a head/eye platform using a variable state dimension filter. Oxford University Engineering Library report number OUEL 1975/93. Submitted to PAMI, 1993.Google Scholar
- 8.P.F. McLauchlan, I.D. Reid, and D.W. Murray. Recursive structure and motion from image sequences. Technical Report OUEL report, in preparation, Dept. Engineering Science, University of Oxford, 1994.Google Scholar
- 9.J. L. Mundy and A. P. Zisserman, editors. Geometric Invariance in Computer Vision. MIT Press, Cambridge, MA, 1992.Google Scholar
- 10.D. W. Murray, P. F. McLauchlan, I. D. Reid, and P. M. Sharkey. Reactions to peripheral image motion using a head/eye platform. In Proc. 4th Int'l Conf. on Computer Vision, Berlin, pages 403–411, Los Alamitos, CA, 1993. IEEE Computer Society Press.Google Scholar
- 11.W.H. Press, B.P. Flannery, S.A. Teukolsky, and W.T. Vetterling. Numerical Recipes in C. Cambridge University Press, 1988.Google Scholar
- 12.L. Quan and R. Mohr. Towards structure from motion for linear features through reference points. In Proc. IEEE Workshop on Visual Motion, 1991.Google Scholar
- 13.I. D. Reid and D. W. Murray. Tracking foveated corner clusters using affine structure. In Proc. 4th Int'l Conf. on Computer Vision, Berlin, pages 76–83, Los Alamitos, CA, 1993. IEEE Computer Society Press.Google Scholar
- 14.P. M. Sharkey, D. W. Murray, S. Vandevelde, I. D. Reid, and P. F. McLauchlan. A modular head/eye platform for real-time reactive vision. Mechatronics, 3(4):517–535, 1993.Google Scholar
- 15.R. Szeliski and S.B. Kang. Recovering 3D shape and motion from image streams using non-linear least squares. Technical Report CRL 93/3, DEC Cambridge Research Lab, March 1993.Google Scholar
- 16.C. Tomasi and T. Kanade. Shape and motion from image streams under orthography: A factorization approach. International Journal of Computer Vision, 9(2):137–154, 1992.Google Scholar
- 17.S. Vinther and R. Cipolla. Towards 3D object model acquisition and recognition using 3D affine invariants. In J. Illingworth, editor, Proc. 4th British Machine Vision Conf., Guildford. BMVA Press, 1993.Google Scholar
- 18.D. Weinshall and C. Tomasi. Linear and incremental acquisition of invariant shape models from image sequences. In Proc. 4th Int'l Conf. on Computer Vision, Berlin, pages 675–682, Los Alamitos, CA, 1993. IEEE Computer Society Press.Google Scholar
- 19.J. Weng, N. Ahuja, and T. S. Huang. Optimal motion and structure estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15(9):864–884, September 1993.Google Scholar