Robust Ellipse-Specific Fitting for Real-Time Machine Vision

This paper presents a robust and non-iterative algorithm for the least-square fitting of ellipses to scattered data. In this work, we undertake a critical analysis of a previous reported work [1] and we propose a novel approach that preserves the advantages while overcomes the major limitations and drawbacks. The modest increase of the computational burden introduced by this method is justified by the achievement of an excellent numerical stability. Furthermore the method is simple and accurate and can be implemented with fixed time of computation. These characteristics coupled to its robustness and specificity makes the algorithm well-suited for applications requiring real-time machine vision.