An algorithm for modeling a set of unordered two-dimensional points by line segments is presented. The points are modeled by highly eccentric ellipses, and line segments are extracted by the major axes of these elongated ellipses. At first, a single ellipse is fitted to points which is then iteratively split to a large number of highly eccentric ellipses to cover the set of points. Then, a merge process follows in order to combine neighboring ellipses with almost collinear major axes to reduce the complexity of the model. Experimental results on various databases show that the proposed scheme is an efficient technique for modeling unordered sets of points and shapes by line segments. A computer vision application of the method is also presented regarding the detection of retinal fundus image features, such as end-points, junctions, and crossovers.