基于改进的K-均值聚类图像分割算法

K-均值聚类是一种被广泛应用的方法。本文提出了基于K-均值聚类的改进算法,并应用于图像分割。针对K-均值聚类算法对离群点的反应过强的缺点,通过替换中心点,比较代价函数,来达到改进划分结果的目的。实验结果表明,该方法能有效改善聚类中心,提高分类精度和准确性。

Image Segmentation Based-on an Improved K-Means Clustering Algorithm

K-Means clustering is a very popular clustering technique which is widely used in numerous applications. This paper presents an improved algorithm for K-Means, and applied it in image segmentation. For the disadvantage of K-Means to the outlier, we improved the result on replacing the centroids and comparing the criterion function. An inspection of the results shows that this method significantly outperformed the K-Means in image segmentation.