Geometric symmetry using rotational scanning method for circular form detection

In this paper, we propose a method to recognize a circular form using a geometric symmetry through a rotational scanning process. Circle recognition is a basic and very important technology in computer vision. It has been continuously studied to improve accuracy and speed as machine vision systems are larger and larger. Circle Hough transform (CHT) is the well-known method for circle detection. However, it requires very huge memory and computing devices to handle data in circle detection. Also, it is difficult to recognize circles in real images when the perspective distortion exists. In this paper, we propose an efficient method to find circles in terms of memory and speed by searching centers and radii of circles at once using a rotational scanning method. Experiments were performed to evaluate the proposed method. The proposed method was compared with the gradient-CHT in Open CV, one of the state-of-the-art methods. Experimental results showed that the proposed method is superior to the existing method.