基于快速均值漂移和水平集方法的图像分割技术

通过加强边缘检测函数的约束作用,对L i提出的无需重新初始化的水平集方法进行了改进,用于提高水平集方法对弱边缘与低对比度图像的分割处理性能。同时,利用快速均值漂移法对图像进行预分割,将得到的轮廓作为水平集方法的初始轮廓,从而降低水平集函数的迭代次数。脑部M R图像的分割实验结果表明,该算法能够准确地提取相应的脑组织,同时大大地提高水平集函数的演化速度。

Fast Mean Shift and Level Set Method for Image Segmentation

By increasing the constrained action of the edge indicator function,the level set evolution without re-initialization is improved for enhancing segmentation of the level set method for images,such as weak edges or low contrast.Original contours of the level set method are obtained by pre-segmentation of the fast mean shift method to decrease iteration times of the level set function.Results of the brain MR image segmentation show that the algorithm can exactly extract the corresponding tissues of the brain and improve the evolution of the level set function.