Abstract
Computer vision is a complex and multidisciplinary field that is still in its early stages of development. Many of the current applications of computer vision are in the field of inspection because of the relatively low cost of these systems and the need for complete inspection in certain high-volume automated manufacturing or assembly processes. The techniques used to inspect the geometric shape and size of an object appear to be application-specific. This article is concerned with the development of a practical method to measure and inspect parts with an elliptical profile. The method is based on the minimization of the sum of the squares of the central norm errors of a given set of boundary points. The central norm error is defined as the Euclidean distance between a boundary point and the point in the best-fit ellipse along the extension line from the center to the boundary point.