Comparative study of Hough Transform methods for circle finding

A variety of circle detection methods which are based on variations of the Hough Transform are investigated. The five methods considered are the standard Hough Transform, the Fast Hough Transform of Li et al.1, a two stage Hough method, and two space saving approaches based on the method devised by Gerig and Klein2. The performance of each of the methods has been compared on synthetic imagery and real images from a metallurgical application. Figures and comments are presented concerning the accuracy, reliability, computational efficiency and storage requirements of each of the methods.