Accurate estimation of elliptical shape parameters from agrey-level image

摘要

The main sources of errors that cause distortion in the shape ofan ellipse are presented, and methods for compensation of thesedistortion factors are formulated. A method for estimation of ellipsecenter coordinates is proposed. A method for the determination of anoptimal elliptical fit to a set of input-edge-points is presented. Anerror function, based on a new geometrical interpretation of an existingerror function, for estimation of all five parameters of an ellipse isdefined, and an objective and independent measure for the goodness ofvarious fits is given. The proposed error function and two otherpreviously developed error functions are applied to six differentsituations. The results of the comparative study show that when all theedge points are used for the fit, the performance of the proposed errorfunction is better than the other two