It applies weighted K-means clustering for region segmentation, instead of traditional K-means clustering.
对于区域分割,使用基于加权平方欧式距离的均值聚类算法代替传统的均值聚类算法。
Firstly, on bisecting K-means is used to quantize image roughly and then we refine the image by improved spectral clustering based weighted distance.
首先利用高效的二分K均值聚类进行粗略量化,然后使用基于加权距离的谱聚类进行再次量化。
A new weighted hybrid C-means clustering based on the K-nearest-neighbour rule is presented in this paper.
该文提出了一种基于K近邻加权的混合C均值聚类算法。
And then based on the K-nearest-neighbour rule, the weighted matrix of samples is computed. Lastly, weighted hybrid C-means clustering based on the K-nearest-neighbour rule is presented.
然后以K近邻规则为基础,计算出样本集的加权矩阵,最后得到基于K近邻加权的混合C均值聚类算法。
And then based on the K-nearest-neighbour rule, the weighted matrix of samples is computed. Lastly, weighted hybrid C-means clustering based on the K-nearest-neighbour rule is presented.
然后以K近邻规则为基础,计算出样本集的加权矩阵,最后得到基于K近邻加权的混合C均值聚类算法。
应用推荐