How can I change each pixel value in a cluster using k-means algorithm to different colors

I am using k-means for segmentation and I have five clusters. I analyze each clusters and find the mean value of each cluster. I then sort each cluster based on the mean value in ascending order. What I want to do is make each pixel value associated with a cluster a different color (e.g. cluster1 = 'red', cluster2 = 'green', cluster3 = 'brown', cluster4 = 'cyan', cluster5 = 'yellow'

Answer 1

[Walter Roberson]

Each pixel will resulted in a cluster index. reshape() that data into rows and columns. image() the result. Then

cmap = [1 0 0; 0 1 0; 0.5 0.25 0.07; 0 1 1; 1 1 0];
colormap(cmap)

[Frank]

Thanks Walter for your response. I would like always for the cluster with the lowest mean to be the color red. In other words, suppose I run kmean algorithm and my cluster index is as follows idx = [3 3 1 2 3;3 1 1 2 3;1 1 2 3 3;2 2 3 3 1;2 3 3 1 2]; And supposed out of this that cluster 3 has the pixels associated with it the lowest mean value. I want to make this cluster "red". I guess what I am saying is that I would like to make whatever cluster has the lowest mean value, I want to make it a particular color each time I run it. I hope I am making sense.

[Walter Roberson]

Calculate the means per cluster, sort the result getting out a sort index, index that sort index at the array of cluster indices, image() that result.

[Frank]

I do calculate the means per cluster and I sort the clusters in ascending order, however I am not sure what you mean by "index that sort index at the array of cluster indices". Could you please clarify?

[Walter Roberson]

[sorted_means, sort_idx] = sort(cluster_means);
remapped_idx = sort_idx(idx);
image(remapped_idx);
colormap(cmap);