NOTE Assessing Error of Fit Functions for Ellipses
Paul L. Rosin · 1996
To avoid the complexity of evaluating the true EuclideanThe fitting of ellipses to edge data is a common task in distance it is usual practice to approximate it by somecomputer vision. In particular, this often arises in the con- measure—the error of fit (EOF) function—that is simplertext of industrial inspection since circular parts in the scene to calculate. The simplest approximation is the algebraicare projected into the image as ellipses. There are many distance [2] calculated from