Is it possible to numerically label a segmented image?

I've written some code to segment, count, and analyze the properties of bacterial colonies in a petri dish. Now, I would like to use principal component analysis to visualize the diversity of bacterial strains present in a given sample. My current method yields the PCA and plots 2 components against each other and even allows me to identify outlier points by number; however, I currently have no way of knowing which colony in my image actually corresponds to number "112" for example. Does anyone know of a way to overlay the colony numbers onto the original image so that I can have a reference point to make sense of the PCA data? I've attached my code if that's helpful.

Thank you in advance!

[Adam Danz]

Hello Katalin, if you have the x,y coordinates of the colonies, you can use labelpoints.m found in the file exchange. This function labels large groups of points like this:

labelpoints(x_vector, y_vector, labels)

and it has an optional outlier parameter that will only label the outliers. But this won't be helpful if you don't have vectors of X,Y values.

Answer 1

[Katalin]

Hello! Try this:

M = im2uint8(label/NUM);
imshow(M)

s = regionprops(M, 'Centroid');

hold on
for k = 1:numel(s)
    c = s(k).Centroid;
    text(c(1), c(2), sprintf('%d', k), ...
        'HorizontalAlignment', 'center', ...
        'VerticalAlignment', 'middle', 'Color', 'red', 'fontsize',12);
end
hold off

[Diondra]

Thank you so much!!