基于灰度直方图的快速K—MEAN图象分割算法

K-MEAN图象分割算法本质上是一种迭代运算,分割结果虽不受初始类中心的影响,但分割处理速度明显依赖于初始类中心的选择.对此,本文根据K-MEAN图象分割算法的原理,提出了基于灰度直方图的快速K-MEAN图象分割算法,该算法直接在灰度直方图上进行迭代运算,不仅减少了数据处理量,且无需人工确定初始类中心,图象分割只需简单的门限判别.理论分析和实验结果证实了该算法能明显加快迭代过程和提高处理速度.

THE FAST K-MEAN IMAGE SEGMENTATION ALGORITHM BASED ON THE GREY HISTOGRAM

A fast k- mean image segmentation algorithm based on grey histogram is presented nere to overcome the iterative procedure of K-MEAN segmentation relying on its selection of original class centers. Iterative operation of this fast K-MEAN segmentation not only decreases the processed data but also to be unecessary to determine its original class centers artificially, and the image segmentation only need simple threshold discrimination. Theoretical analysis and experimental results show that it can quicken the iterative procedure and decrease processing time greatly.