Indent round detection of Brinell test images based on Snake model and shape similarity

In this paper a new indent round detection scheme of Brinell test images is proposed. An improved algorithm based on Snake model and shape similarity is presented, which is able to change the parameters of Snake model adaptively according to the shape similarity between the Snake curve and the prior shape information. Effective contour extraction is processed with this scheme, and indent round is successfully detected employing Hough transform. Various experiment results for synthesized and real images show that this scheme is promising.