How can i choose the k initial centroids far away from each other in k-means clustering based image segmentation

The steps performed for k-means clustering are as follows:

1. Choose k initial centroids
2. Compute the distance from each pixel to the centroid
3. Recalculate the centroids after all the pixels have been assigned
4. Repeat steps 2 and 3 until the same points are assigned to each cluster in consecutive rounds.

How can i choose the k-initial centroids, such that they are far from each other.

Answer 1

[Alok Nimrani]

You can make use of k-means++ algorithm to choose the initial centroids far away from each other. This algorithm is the one used by default while performing k-means clustering using the k-means function in MATLAB.

Hope this helps.