Abstract
We investigate several numerical schemes for estimating parameters in computer vision problems: HEIV, FNS, renormalization method, and others. We prove mathematically that these algorithms converge rapidly, provided the noise is small. In fact, in just 1-2 iterations they achieve maximum possible statistical accuracy. Our results are supported by a numerical experiment. We also discuss the performance of these algorithms when the noise increases and/or outliers are present.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
N. Chernov and C. Lesort, “Statistical efficiency of curve fitting algorithms,” Computational Statistics and Data Analysis, Vol. 47, pp. 713–728, 2004.
W. Chojnacki, M.J. Brooks, and A. van den Hengel, “Rationalising the renormalisation method of Kanatani,” Journal of Mathematical Imaging and Vision, Vol. 14, pp. 21–38, 2001.
W. Chojnacki, M.J. Brooks, A. van den Hengel, and D. Gawley, “From FNS to HEIV: A link between two vision parameter estimation methods,” EEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 26, pp. 264–268, 2004.
W. Chojnacki, M.J. Brooks, A. van den Hengel, and D. Gawley, “FNS, CFNS and HEIV: A unifying approach,” Journal of Mathematical Imaging and Vision, Vol. 23, pp. 175–183, 2005.
K. Kanatani, “Statistical bias of conic fitting and renormalization,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 16, pp. 320–326, 1994.
K. Kanatani, Statistical Optimization for Geometric Computation: Theory and Practice, Elsevier: Amsterdam, 1996.
K. Kanatani, “Cramer-Rao lower bounds for curve fitting,” Graphical Models and Image Processing, Vol. 60, pp. 93–99, 1998.
K. Kanatani, “Uncertainty modeling and model selection for geometric inference,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 26, pp. 1307–1319, 2004.
K. Kanatani, “For geometric inference from images, what kind of statistical model is necessary?” Systems and Computers in Japan, Vol. 35, pp. 1–9, 2004.
K. Kanatani, “Further improving geometric fitting,” in Proceedings 5th International Conference on 3-D Digital Imaging Modeling, Ottawa, Canada, 2005, pp. 2–13.
K. Kanatani, “Hyperaccuracy for geometric fitting,” in 4th Int. Workshop Total Least Squares and Errors-in-Variables Modelling, Leuven, Belgium, August 2006.
Y. Leedan, Statistical analysis of quadratic problems in computer vision, Ph.D. thesis, ECE Department, Rutgers Univ., May 1997.
Y. Leedan and P. Meer, “Heteroscedastic regression in computer vision: Problems with bilinear constraint,” Intern. J. Comp. Vision, Vol. 37, pp. 127–150, 2000.
B. Matei and P. Meer, “A General Method for Errors-in-Variables Problems in Computer Vision,” in Proceedings, CVPR 2000, IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Hilton Head Island, South Carolina, 2000, IEEE Computer Society Press: Los Alamitos, CA 2000, Vol. 2, pp. 18–25.
B. Matei and P. Meer, “Reduction of Bias in Maximum Likelihood Ellipse Fitting,” in Proceedings, 15th International Conference on Computer Vision and Pattern Recognition, Barcelona, Spain, 2000, Vol. 3, pp. 802–806.
G. Taubin, “Estimation of Planar Curves, Surfaces and Nonplanar Space Curves Defined by Implicit Equations, with Applications to Edge and Range Image Segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 13, pp. 1115–1138, 1991.
Author information
Authors and Affiliations
Corresponding author
Additional information
Nikolai Chernov PhD in mathematics from Moscow University in 1984. Researcher in JINR (Dubna, Russia) in 1984–91. Professor of Mathematics at University of Alabama at Birmingham, USA, since 1994.
Rights and permissions
About this article
Cite this article
Chernov, N. On the Convergence of Fitting Algorithms in Computer Vision. J Math Imaging Vis 27, 231–239 (2007). https://doi.org/10.1007/s10851-006-0646-1
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10851-006-0646-1