催化剂分割算法的初步研究

目的：在TS—MRF模型基础上，针对基于二叉树分割算法的分割停止判断提出新的算法。方法：在分割图像中引入试探图像作为“催化剂”，并以“催化剂”在分割结果中是否显现为依据来判断二叉树分割的候选结点是否应该停止分割，具体分割算法采用MRF分割。结果：对多组模拟图像进行了初步分割实验，实验表明该方法可以对实验图像的树状分割过程给出正确的停止判断。结论：催化剂分割算法可有效地对待分割图像进行正确的类数判断。

Preliminary Study of Catalyst Segmentation Method

Objective： propose a novel method to judge whether a segmentation step ought to stop or not. Method： we import one image as the catalyst into segmentation algorithm, and make some rules to decide the candidate nodes whether stopping segmentation or not. Result： Primary experiments show that CS is simple and useful. Conclusion： The correct classes of the image can be obtained by catalyst segmentation method efficiently.