基于图像森林变换算法的区域增长方法研究

Research on Region Rising with the Image Foresting Transform

引入了一种基于图的连通性的图像分割算法———图像森林变换算法.通过对其特性的仿真和分析,在目标区域内灰度级变化范围不大的情况下,提出了一种提高执行速度、减少存储空间的IFT新算法,仿真结果证明了该算法的可行性.

The image foresting transform（IFT） , which is a graph-based algorithm for design of image processing operators based on connectivity, is researched. Characteristics of IFT algorithm are analyzed by emulations. In some condition that the change of the object＇s gray scale is relatively uniform, an improved algorithm which has faster speed and needs less memory spaces is presented. The feasibility of the improved algorithm is proved by emulations.