An Image Segmentation Method Based on the Improved Snake Model

Getting the contour of the object based on the snake model is an important method in the image segmentation. In this paper the authors first introduce the theory of the traditional snake model. A new snake model based on a dot which is in the object is been proposed for avoiding some drawbacks of the traditional snake model. The algorithm not only inherits the topology ability of the traditional snake, but also has the ability of convergence to the concave, and the convergent rate is also added. The segmentation effort of the algorithm is proved by the experiments.