Clever Geek Handbook

Image segmentation using K means

When I execute the following command in Matlab 2012a

centroids=kmeans(imread('image.jpg'),4);

I am getting the following error:

Error using  +
Integers can only be combined with integers of the same class, or scalar doubles.

Error in kmeans>distfun (line 659)

            D(:,i) = D(:,i) + (X(:,j) - C(i,j)).^2;

Error in kmeans (line 273)

D = distfun(X, C, distance, 0, rep, reps);

I need to segment an image into 4 clusters. The image is an image of a brain tumor in a brain format JPEG

. The size of this image 233x216

.
Please give me a solution to group this image file.

+1


source to share


2 answers


Use the kmeans segmentation algorithm instead of the standard kmeans algorithm provided in MATLAB.



Refer to this file . This is the K algorithm used for segmentation purposes. Using this algorithm, my program works.

+4


source


Color image (MxNx3) may be causing the problem

If you want to copy the intensity values ​​of an image into 4 clusters, you must do

im = imread('image.jpg');
im=rgb2gray(im) //if you only want grayscale intensities
[idx centroids]=kmeans(double(im(:)),4);

If you want to consider a color, you can do something like



im = imread('image.jpg'); 
im = reshape(im,size(im,1)*size(im,2),size(im,3))
[idx centroids]=kmeans(double(im(:)),4);

To visualize segmentation, you can do something like

imseg = zeros(size(im,1),size(im,2));
for i=1:max(idx)
    imseg(idx==i)=i;
end
imagesc(imseg)
+4


source






More articles:


All Articles