Contents

• Introduction
• A Glance over Parameter Estimation in General
• Conic Fitting Problem
• Least-Squares Fitting Based on Algebraic Distances
  • Normalization with A+C=1
  • Normalization with
  • Normalization with F=1
• Least-Squares Fitting Based on Euclidean Distances
  • Why are algebraic distances usually not satisfactory ?
  • Orthogonal distance fitting
• Gradient Weighted Least-Squares Fitting
• Bias-Corrected Renormalization Fitting
• Kalman Filtering Technique
  • Standard Kalman Filter
  • Extended Kalman Filter
  • Discussion
  • Iterated Extended Kalman Filter
  • Application to Conic Fitting
• Robust Estimation
  • Introduction
  • Clustering or Hough Transform
  • Regression Diagnostics
  • M-estimators
  • Least Median of Squares
• Two Examples
  • Noisy data without outliers
  • Noisy data with outliers
• Conclusions
• References
• About this document ...