Further Improving Geometric Fitting
Author:
- 1 June 2005
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15506185,p. 2-13
Abstract
We give a formal definition of geometric fitting in a way that suits computer vision applications. We point out that the performance of geometric fitting should be evaluated in the limit of small noise rather than in the limit of a large number of data as recommended in the statistical literature. Taking the KCR lower bound as an optimality requirement and focusing on the linearized constraint case, we compare the accuracy of Kanatani's renormalization with maximum likelihood (ML) approaches including the FNS of Chojnacki et al. and the HEIV of Leedan and Meer. Our analysis reveals the existence of a method superior to all these.Keywords
This publication has 13 references indexed in Scilit:
- Statistical efficiency of curve fitting algorithmsComputational Statistics & Data Analysis, 2004
- Uncertainty modeling and model selection for geometric inferenceIEEE Transactions on Pattern Analysis and Machine Intelligence, 2004
- From fns to heiv: a link between two vision parameter estimation methodsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2004
- For geometric inference from images, what kind of statistical model is necessary?Systems and Computers in Japan, 2004
- Unbiased Errors-In-Variables Estimation Using Generalized Eigensystem Analysis2004
- Toward a statistically optimal method for estimating geometric relations from noisy data: cases of linear relations2003
- Comparing optimal three-dimensional reconstruction for finite motion and optical flowJournal of Electronic Imaging, 2003
- Heteroscedastic Regression in Computer Vision: Problems with Bilinear ConstraintInternational Journal of Computer Vision, 2000
- Cramer–Rao Lower Bounds for Curve FittingGraphical Models and Image Processing, 1998
- Information Geometry of Estimating Functions in Semi-Parametric Statistical ModelsBernoulli, 1997