Clustering coordinates on image

How would i cluster these points so that i am left wth 4 sets of coordinates in a 4 by 2 matrix currently it returns a 13 by 2.

This is the returned matrix.

centers =

   16.5000    9.0000
    8.5000   29.0000
  423.0000   29.0000
   18.5000   33.0000
   30.0000   35.0000
  448.5000   35.0000
  431.5000   36.5000
   37.5000   45.5000
  456.5000  435.0000
   43.5000  450.0000
  426.5000  450.0000
  445.0000  457.5000
  439.5000  469.0000

Answer 1

[Image Analyst]

Try this:

centers =[...
	16.5000    9.0000
	8.5000   29.0000
	423.0000   29.0000
	18.5000   33.0000
	30.0000   35.0000
	448.5000   35.0000
	431.5000   36.5000
	37.5000   45.5000
	456.5000  435.0000
	43.5000  450.0000
	426.5000  450.0000
	445.0000  457.5000
	439.5000  469.0000];
subplot(1, 2, 1);
plot(centers(:, 2), centers(:, 1), 'r+', 'MarkerSize', 9, 'LineWidth', 2);
[indexes, centroids] = kmeans(centers, 4)
axis ij; % Flip axis direction.
grid on;
subplot(1, 2, 2);
gscatter(centers(:, 2), centers(:, 1), indexes);
axis ij; % Flip axis direction.
grid on;
% Put a big circle at the centroid
hold on;
plot(centroids(:, 2), centroids(:, 1), 'bo', 'MarkerSize', 30, 'LineWidth', 2);

You now have 4 clusters.

[jg]

Thank you