基于变分等值线方法的图像分割技术

IMAGE SEGMENTATION BASED ON VARIATIONAL CONTOURS METHOD

使用level set,函数来隐式地追踪目标图像,给出了level set函数的变分模型和相应的 Euler-Lagrange方程．在保证level set函数的等值线在目标图像附近的运动速度渐趋于零的前提下,改进了原来的Euler-Lagrange疗程,加快迭代速度．可灵活地调节模型参数以满足不同的分割要求．文中算例说明本模型的有效性、灵活性．

In this paper, a level set function is used to detect the object implicitly. A variational model of the level set function and the corresponding Euler-Lagrange equation are proposed. The Euler-Lagrange equation is modified to accelerate the convergence rate, Simultaneously, the zero contour of level set function would stop on the boundary of the object and detects it. Finally, the parameters in the model can be selected flexibly to satisfy some different motivations in segmentation. A number of examples show the efficiency and the flexibleness of our new model.