Ellipse detection using a genetic algorithm

Abstract

In this paper we present a new algorithm to detect ellipses in an intensity image. The algorithm first extracts line-support regions from the original image. Next, candidates for elliptic arcs are selected from these line-support regions. In the proposed algorithm, each ellipse is defined by a triplet of line-support regions. And, using a genetic algorithm, the proposed algorithm searches for triplets of line-support regions that have the highest fitness to the image. By experiments using natural images, we estimate the performance of the proposed algorithm.