一种无重新初始化的水平集方法的错觉轮廓捕捉新方法

在错觉轮廓捕捉模型建立前,我们要得到根据物体边界的符号距离函数时,用Eikonal方程不能实现的,我们用基于水平集方法的分割技术实现,扩大了模型的使用范围;在Zhu和Chan等人的错觉轮廓捕捉模型基础上引入了李纯明等人提出的符号距离约束信息,这就使得在水平集函数演化时不必对其重新初始化,并大大简化了模型的数值处理水平集函数的演化速度.并通过实验验证了该方法的优势.

A Method of Capturing'Illusory Contour Based on Level Set Method without Re-initialization

Before buiding the model of capturing illusory contour,it is necessary to get the signed distance function according to the edges of the objects.To the images which cannot he processed hy Eilonal equation, we use the technique of image segmentation based on level set method to get the signed distance function we want.And then we extend the illusory contour model to a more wide range.We add the constrained information of the signed distance function to the model proposed,by Zhu and Chan,by which we need no re- initialization of the level set function during its evolution.With the constraint,we can make the numerical analysis much easier and the level set evolutes much faster.Numerical experiments show the efficiency.