A New Thinning Algorithm for Removing Noise-Spurs and Retaining End-Points
Jun S. Huang and Wen J. Lee Institute of Information Science, Academia Sinica, Taipei, Taiwan, Republic of China
A new algorithm for thinning a figure in a picture is developed. The main consideration here is the identification of end-points and noise-spurs. We discuss this in detail as five patterns of an end-point (or noise-spur) emerge in our study. The experimental results show that the new algorithm is better than the Davies and Plummer's method [6]. Especially when shapes are subjected to noise the new method performs almost as the ideal cases.