Image segmentation using gaussian mixture model clustering based on the blobworld paper

I'm new to machine learning implementation and I'm trying to implement the image segmentation approach from the Blobworld paper. I have extracted the 6 feature vectors of the image that are described in the paper and I have them saved in matrix X which has 6 rows that are each a feature and number of pixels from the input image as columns.

I have managed to write the code that extracts the features. and now I need to cluster the feature vectors into gaussian clusters using EM:

 X = double(computeBlobworldFeatureVectors(original_image));%original_image is the image from imread('zebra.jpg'); same image from the paper
 X = X([1 2 3 6 5 4], :);%l*,a*,b*,contrast,anistropy,polarity
 

I am now trying to do (d) from the following image that is snapped from the paper :

Answer 1

[Asvin Kumar]

Here's a popular implementation of the EM algorithm from File Exchange which might be of use to you:

Expectation Maximization Algorithm by Reza Ahmadzadeh.

----------

This is one of the several submissions in MATLAB File Exchange on MATLAB Central which is a forum for our product users to interact, exchange information and knowledge, without MathWorks' involvement. Feel free to contact the author of this submission directly for specific questions about the implementation