各位先進及各位高手, 小弟在IEEE上下載一篇有關於影像分割的Journal " Contrast enhancement based on divided histogram manipulation" 其中有一段k-mean將影像分求得其質量中心,歸類成k個群組 想請教各位先進與高手,能提供小弟方向或程式供參考用! 感激不盡!謝謝! 以下為k-mean 演算法 The recursive version of the K-mean Algorithm: (1)Start out with k-clusters, which initialized with one sample data picked at random. Find the centroid closest to each of the n-2 samples remained. Place the sample in the cluster with the closest centroid identified. After having each sample assigned, re-compute the centroid of the cluster changed. (2)Revisit the data again .Find for each sample the centroid closest to it. Place the sample in the cluster with the closest centroid identified (In this step, there is no need for re-computing of any centroid.) (3)Terminate if no cluster changes its center. (4) Re-compute the centroids of clusters and repeat step 2.