最大流方法在图像去噪和分割中的研究进展

Progress of Max-Flow Method in Image Denoising and Segmentation

近年来,基于能量泛函的图像去噪和分割模型得到广泛的关注,解决这类模型的有力工具之一是最大流方法。本文分别介绍离散的和连续的最大流方法,包括能量泛函最小化问题转化为最大流问题的基本步骤,以及相应最大流问题的求解方法,并展望最大流方法的发展前景。

In recent years, image denoising and segmentation model based on energy functional received widely atten- tion. Max-flow method is one of the most powerful tools to solve this kind of model. Discrete and continuous max-flow methods are introduced, they both include the basic steps in which energy functional minimization problem transformed to max-flow problem, and the solutions of the corresponding max-flow problem are also reviewed. In addition, the de- velopment of max-flow method are also discussed.