摘要： This paper proposes a novel active contour model for image segmentation based on techniques of curve evolution. The paper introduces an energy functional including a local fuzzy energy and a global fuzzy energy to attract the active contour and stop it on the object boundaries. The local term allows the method to deal with intensity inhomogeneity in images. The global term, aside from driving the contour toward the desired objects, is used to avoid unsatisfying results led by unsuitable initial contour position, a common limitation of models using local information solely. In addition, instead of solving the Euler–Lagrange equation, the paper directly calculates the alterations of the fuzzy energy. By this way, the contour converges quickly to the object boundary. Experimental results on both 2D and 3D images validate the effectiveness of the model when working with intensity inhomogeneous images. 其他題名： Nonlinear Dyn 出版者： Dordrecht: Springer Netherlands 出版日期： 2012-01 出處： Nonlinear dynamics, 2012-01, Vol.67 (2), p.1559-1578 版權： Springer Science+Business Media B.V. 2011 版權： Nonlinear Dynamics is a copyright of Springer, (2011). All Rights Reserved. 識別號： ISSN: 0924-090X 識別號： EISSN: 1573-269X 識別號： DOI: 10.1007/s11071-011-0088-1
