Detect brain tumor using Color based KMeans clustering

Can anyone tell me why am i getting warning of empty cluster created at some replications. Moreover the demo at Matlab: http://www.mathworks.com/products/image/examples.html?file=/products/demos/shipping/images/ipexhistology.html uses Lab colorspace but i want to implement using rgb values, how can i do this

4 Comments

While performing k-means I get error message in the line
lab_he=applycform(d,cform)
>>Error in ==> applycformsequence at 12
out = applycform(out, cforms{k});
Error in ==> applycform at 86
out = c.c_func(columndata, cdata{:});
Error in ==> applycformsequence at 12
out = applycform(out, cforms{k});
Error in ==> applycform at 86
out = c.c_func(columndata, cdata{:});
Error in ==> mammu at 35
lab_he = applycform(d,cform);
>>
There would have been at least one line of error message proceeding that. Something like "recursion limit of 500 reached" perhaps.

Sign in to comment.

Answers (1)

The demo starts with an RGB image, then converts to lab color space. You can leave it in RGB color space if you want. Just don't do the conversion.
I'm not sure why your brain tumor image is in color? What was the source of it? CT, MRI? Why isn't it monochrome?

2 Comments

i have converted the grayscale image to rgb.
You used false coloring? Or you converted each grayscale array to one of its RGB equivalents by copying the information to all three color planes?

Sign in to comment.

Asked:

on 5 Jul 2012

Commented:

on 17 Oct 2016

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!