Very fast best-fit circular and elliptical boundaries by chord data

Area of Science:

• Computer Vision
• Image Processing
• Geometric Modeling

Background:

• Circle and ellipse fitting are crucial for delineating objects in machine vision.
• Algebraic least-squares methods are computationally efficient but prone to bias and outliers.

Purpose of the Study:

• To introduce a novel real-time, least-squares method for fitting elliptic primitives.
• To address the limitations of existing algebraic methods in terms of bias and outlier sensitivity.

Main Methods:

• An indirect geometric fit using the quadratic polynomial form of parallel chord lengths.
• A real-time, least-squares approach for ellipse fitting in noisy image data.

Main Results:

• The proposed method demonstrates superior computational efficiency over algebraic techniques.
• The algorithm shows enhanced robustness to outlier data.
• Experimental results indicate reduced bias error in parameter estimation.

Conclusions:

• The novel indirect geometric fitting method offers a more accurate and robust solution for ellipse fitting in machine vision.
• This approach is suitable for real-time applications requiring reliable object delineation.