Matlab Code SImple LVQ / SOM

function scan(img)

files = dir('*.jpg');

hist = [];

for n = 1 : length(files)

    filename = files(n).name;

    file = imread(filename);

 

    hist = [hist, imhist(rgb2gray(imresize(file,[ 50 50])))]; %#ok

end

 

som = selforgmap([10 10]);

som = train(som, hist);

t   = som(hist); %extract class data

 

net = lvqnet(10);

net = train(net, hist, t);

 

like(img, hist, files, net)

end




====================




function like(im, hist, files , net)

    hs = imhist(rgb2gray(imresize(im,[50 50])));

    cls = vec2ind(net(hs));

 

    [~, n] = size(hist);

    for i = 1 : n

        if(cls == vec2ind(net(hist(:, i))))

            figure('name', files(i).name);

            imshow(imresize(imread(files(i).name), [100 100]))

        end

    end

end
















Reff:

https://scriptbucket.wordpress.com/2012/09/21/image-classification-using-matlab-somlvq/