基于图割的图像分割方法及其新进展

The Basic Principle and Its New Advances of Image Segmentation Methods Based on Graph Cuts

鉴于图割的理论意义和实际应用价值,系统综述了基于图割的图像分割方法.首先,深入分析了基于图割的图像分割方法的基本原理,主要从定性和定量角度剖析了图割与能量函数最小化之间的关系,然后,概括了基于图割的图像分割方法的基本步骤,包括能量函数的设计、图的构造和最小割/最大流方法,其次,系统梳理和评述了基于图割的图像分割方法的国内外研究现状,最后,指出了基于图割的图像分割方法的发展方向.

In view of the theoretical significance and practical value of graph cuts, the image segmentation methods based on graph cuts are reviewed in this paper. Firstly, the basic principle of image segmentation method based on graph cuts is analyzed in detail, which mainly focuses on the relation between graph cuts and energy minimization involving both qualitative and quantitative analysis. Secondly, the steps of image segmentation methods based on graph cuts are generalized as designing energy function, constructing graph, and minimum cut/maximum flow approaches. Thirdly, the current status of image segmentation methods based on graph cuts is combed and commented. Finally, the future for these segmentation methods is pointed out.