S. Anam, E. Uchino, N. Suetake
The Level Set Method has been widely used in image processing and computer vision. However, the Level Set Method has problems with stability and convergence. This arrangement often leads to either a complete breakdown or premature termination of the curve evolution process, the consequence of which is unsatisfactory results. Convergence and stability of the Level Set Method are controlled by the speed function and parameters of the Level Set Method. In this paper, we will present a modification of the speed function for image boundary detection. To preserve the boundary of an image and to remove noise, we use a diffusion filter to substitute a Gaussian Filter in the standard Level Set Method. After evaluating the experiments using a specific set of artificial images, we have found that the performance of those methods has been significantly improved. © 2013 The Authors.
Graduate School of Science and Engineering, Yamaguchi University, Yamaguchi 753-8512, 1677-1 Yoshida, Japan; Department of Mathematics, University of Brawijaya, Malang 65145, Indonesia; Fuzzy Logic Systems Institute, Iizuka, Fukuka 820-0067, 680-41 Kawazu, Japan
