A Comparison of Ellipse Fitting Methods and Implications for Multiple-View Geometry Estimation

After summarising the conceptual relationship between some old and new techniques for fitting ellipses to data, we conduct a thorough experimental comparison of the discussed methods. The newer techniques promise consistency, unbiasedness, or hyper-accuracy, but our experiments reveal that in practi...

Full description

Saved in:
Bibliographic Details
Published in2012 International Conference on Digital Image Computing: Techniques and Applications pp. 1 - 8
Main Authors Szpak, Z. L., Chojnacki, W., van den Hengel, A.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.12.2012
Subjects
Approximation methods
Cost function
Equations
Maximum likelihood estimation
Shape
Training
Online AccessGet full text
ISBN9781467321808
146732180X
DOI10.1109/DICTA.2012.6411722

Cover

More Information
Summary:After summarising the conceptual relationship between some old and new techniques for fitting ellipses to data, we conduct a thorough experimental comparison of the discussed methods. The newer techniques promise consistency, unbiasedness, or hyper-accuracy, but our experiments reveal that in practice the performance of these estimators is difficult to distinguish from the performance of established estimators. Since the newer techniques could potentially be extended to other estimation problems, we discuss what implications our experimental findings have for solving more general multiple-view geometry estimation problems.
ISBN:9781467321808
146732180X
DOI:10.1109/DICTA.2012.6411722