International Journal of Computer Vision

, Volume 80, Issue 2, pp 167–188 | Cite as

Statistical Optimization for Geometric Fitting: Theoretical Accuracy Bound and High Order Error Analysis

Article

Abstract

A rigorous accuracy analysis is given to various techniques for estimating parameters of geometric models from noisy data. First, it is pointed out that parameter estimation for vision applications is very different in nature from traditional statistical analysis and hence a different mathematical framework is necessary. After a general framework is formulated, typical numerical techniques are selected, and their accuracy is evaluated up to high order terms. As a byproduct, our analysis leads to a “hyperaccurate” method that outperforms existing methods.

Keywords

Geometric fitting Parameter estimation Error analysis Hyperaccuracy KCR lower bound 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Akaike, H. (1977). A new look at the statistical model identification. IEEE Transactions on Automatic Control, 16(6), 716–723. MathSciNetGoogle Scholar
  2. Amari, S., & Kawanabe, M. (1997). Information geometry of estimating functions in semiparametric statistical models. Bernoulli, 3, 29–54. MATHCrossRefMathSciNetGoogle Scholar
  3. Begelfor, E., & Werman, M. (2005). How to put probabilities on homographies. IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(10), 1666–1670. CrossRefGoogle Scholar
  4. Bickel, P. J., Klassen, C. A. J., Ritov, Y., & Wellner, J. A. (1994). Efficient and adaptive estimation for semiparametric models. Baltimore: Johns Hopkins University Press. Google Scholar
  5. Bookstein, F. L. (1979). Fitting conic sections to scattered data. Computer Graphics and Image Processing, 9(1), 56–71. CrossRefGoogle Scholar
  6. Chernov, N., & Lesort, C. (2004). Statistical efficiency of curve fitting algorithms. Computational Statistics and Data Analysis, 47(4), 713–728. CrossRefMathSciNetGoogle Scholar
  7. Chojnacki, W., Brooks, M. J., van den Hengel, A., & Gawley, D. (2000). On the fitting of surfaces to data with covariances. IEEE Transactions on Pattern Analysis and Machine Intelligence, 22(11), 1294–1303. CrossRefGoogle Scholar
  8. Chojnacki, W., Brooks, M. J., & van den Hengel, A. (2001). Rationalising the renormalisation method of Kanatani. Journal of Mathematical Imaging and Vision, 14(1), 21–38. MATHCrossRefMathSciNetGoogle Scholar
  9. Chojnacki, W., Brooks, M. J., van den Hengel, A., & Gawley, D. (2004). From FNS to HEIV: A link between two vision parameter estimation methods. IEEE Transactions on Pattern Analysis and Machine Intelligence, 26(2), 264–268. CrossRefGoogle Scholar
  10. Chojnacki, W., Brooks, M. J., van den Hengel, A., & Gawley, D. (2005). FNS, CFNS and HEIV: A unifying approach. Journal of Mathematical Imaging and Vision, 23(2), 175–183. CrossRefMathSciNetGoogle Scholar
  11. Hartley, R., & Zisserman, A. (2000). Multiple view geometry in computer vision. Cambridge: Cambridge University Press. MATHGoogle Scholar
  12. Kanatani, K. (1993). Renormalization for unbiased estimation. In Proceedings of the 4th international conference on computer vision (pp. 599–606), Berlin, Germany, May 1993. Google Scholar
  13. Kanatani, K. (1996). Statistical optimization for geometric computation: theory and practice. Amsterdam: Elsevier. Reprinted by Dover, New York (2005). MATHCrossRefGoogle Scholar
  14. Kanatani, K. (1998). Cramer–Rao lower bounds for curve fitting. Graphical Models and Image Processing, 60(2), 93–99. CrossRefGoogle Scholar
  15. Kanatani, K. (2004). Uncertainty modeling and model selection for geometric inference. IEEE Transactions on Pattern Analysis and Machine Intelligence, 26(10), 1307–1319. CrossRefGoogle Scholar
  16. Kanatani, K. (2005). Further improving geometric fitting. In Proceedings of the 5th international conference on 3D digital imaging and modeling (pp. 2–13), Ottawa, Canada, June 2005. Google Scholar
  17. Kanatani, K. (2006). Ellipse fitting with hyperaccuracy. IEICE Transactions on Information and Systems, E89-D(10), 2653–2660. CrossRefGoogle Scholar
  18. Kanatani, K., & Ohta, N. (2003). Comparing optimal three-dimensional reconstruction for finite motion and optical flow. Journal of Electronic Imaging, 12(3), 478–488. CrossRefGoogle Scholar
  19. Kanatani, K., & Sugaya, Y. (2000). High accuracy fundamental matrix computation and its performance evaluation. IEICE Transactions on Information and Systems, E90-D(2), 579–585. CrossRefGoogle Scholar
  20. Kanatani, K., & Sugaya, Y. (2007). Performance evaluation of iterative geometric fitting algorithms. Computational Statistics and Data Analysis, 52(2), 1208–1222. CrossRefGoogle Scholar
  21. Kanatani, K., Ohta, N., & Kanazawa, Y. (2000). Optimal homography computation with a reliability measure. IEICE Transactions on Info and Systems, E83-D(7), 1369–1374. Google Scholar
  22. Leedan, Y., & Meer, P. (2000). Heteroscedastic regression in computer vision: Problems with bilinear constraint. International Journal of Computer Vision, 37(2), 127–150. MATHCrossRefGoogle Scholar
  23. Neyman, J., & Scott, E. L. (1948). Consistent estimates based on partially consistent observations. Econometrica, 16(1), 1–32. CrossRefMathSciNetGoogle Scholar
  24. Ohta, N. (2003). Motion parameter estimation from optical flow without nuisance parameters. In 3rd international workshop on statistical and computational theory of vision, Nice, France, October 2003. http://www.stat.ucla.edu/~sczhu/Workshops/SCTV2003.html.
  25. Okatani, T., & Deguchi, K. (2003). Toward a statistically optimal method for estimating geometric relations from noisy data: Cases of linear relations, In Proceedings of the IEEE conference on computer vision and pattern recognition (Vol. 1, pp. 432–439), Madison, WI, USA, June 2003. Google Scholar
  26. Pennec, X., Fillard, P., & Ayache, N. (2006). A Riemannian framework for tensor computing. International Journal of Computer Vision, 66(1), 41–66. CrossRefMathSciNetGoogle Scholar
  27. Rissanen, J. (1989). Stochastic complexity in statistical inquiry. Singapore: World Scientific. MATHGoogle Scholar
  28. Sampson, P. D. (1982). Fitting conic sections to “very scattered” data: An iterative refinement of the Bookstein algorithms. Computer Graphics and Image Processing, 18(1), 97–108. CrossRefGoogle Scholar
  29. Taubin, G. (1991). Estimation of planar curves, surfaces, and non-planar space curves defined by implicit equations with applications to edge and range image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13(11), 1115–1138. CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2007

Authors and Affiliations

  1. 1.Department of Computer ScienceOkayama UniversityOkayamaJapan

Personalised recommendations