I need a matlab code for color reduction using self-organizing map (SOM) learning

Please Help me. I stuck here and need the code urgently. I m using color images which are represented in L* u* v* color space.....

please help me.. URGENT

Answer 1

[Image Analyst]

Why does it need to be SOM? Have you tried Brett's tool: http://www.mathworks.com/matlabcentral/fileexchange/38484-segmenttool-an-interactive-gui-for-segmenting-images Or maybe http://www.mathworks.com/matlabcentral/fileexchange/37197-dem-diffused-expectation-maximisation-for-image-segmentation Or even simply rgb2ind()?

[Purbankan]

Sir I m trying to first reduce the color of the image, and then use simulated annealing for seeking the optimal clusters using the image colors which has been projected into a small set of prototypes using SOM.

Anyway this function rgb2ind(), isnt this function for RGB images? how can i use this function for L*u*v*?

[Image Analyst]

It performs clustering, so it should probably work, though I haven't tried it on anything but uint8 images. If you have Luv images they're floating point probably, and in a different range than 0-255. Perhaps you can convert Luv to RGB if you find out it's necessary. But bottom line, I don't have any SOM code.

Answer 2

[KAE]

Matlab offers self organizing map functions functions but you have to buy the Neural Network toolbox.