分片常值灰度图像的快速分割

摘要

PC 模型是一个著名的基于区域的活动轮廓模型,它实际上是利用水平集方法解决分片常值灰度图像的分割问题.提出一个以偏微分方程形式表达的新模型,它可以看成是 PC 模型的一种改进.实验显示:新模型能够实现分片常值灰度图像的快速分割,同时迭代次数对初始轮廓的大小和位置不敏感.%Piecewise Constant(PC) model is one of the most popular region-based models in the existing active contours. It is actually devoted to the problem of piecewise-constant segmentation of images using the level set approach. This paper proposes a modified version of PC model in a Partial Differential Equation(PDE) formulation. The proposed model can handle the fast segmentation of images with piecewise-constant intensity,moreover,the number of iterations is significantly less sensitive to initialization of the contours.