face_recognition_steps.m

% https://blog.cordiner.net/2010/12/02/eigenfaces-face-recognition-matlab/
% get directories for all training images
files = strcat(pwd, '/ORL_images')
fn = getfn(files, 'pgm$');


% read and convert images to column vector
image_dims = [112, 92];
num_images = numel(fn);
images = [];
for n = 1:num_images
    img = imread(fn{n});
    if n == 1
        images = zeros(prod(image_dims), num_images);
    end
    images(:, n) = img(:);
end


% steps 1 and 2: find the mean image and the mean-shifted input images
mean_face = mean(images, 2);
shifted_images = images - repmat(mean_face, 1, num_images);
 
% steps 3 and 4: calculate the ordered eigenvectors and eigenvalues
[evectors, score, evalues] = pca(images');
 
% step 5: only retain the top 'num_eigenfaces' eigenvectors (i.e. the principal components)
num_eigenfaces = 20;
evectors = evectors(:, 1:num_eigenfaces);
 
% step 6: project the images into the subspace to generate the feature vectors
features = evectors' * shifted_images;



% Classification to test face recognition
% calculate the similarity of the input to each training image
feature_vec = evectors' * (input_image(:) - mean_face);
similarity_score = arrayfun(@(n) 1 / (1 + norm(features(:,n) - feature_vec)), 1:num_images);
 
% find the image with the highest similarity
[match_score, match_ix] = max(similarity_score);
 
% display the result
figure, imshow([input_image reshape(images(:,match_ix), image_dims)]);
title(sprintf('matches %s, score %f', filenames(match_ix).name, match_score));




% display the eigenvectors
figure;
for n = 1:num_eigenfaces
    subplot(2, ceil(num_eigenfaces/2), n);
    evector = reshape(evectors(:,n), image_dims);
    imshow(evector);
end



% display the eigenvalues
normalised_evalues = evalues / sum(evalues);
figure, plot(cumsum(normalised_evalues));
xlabel('No. of eigenvectors'), ylabel('Variance accounted for');
xlim([1 30]), ylim([0 1]), grid on;