very low score similarity in pca

3 visualizaciones (últimos 30 días)

primrose khaleed el 1 de Jun. de 2014

0
Enlazar

Enlace directo a esta pregunta

https://es.mathworks.com/matlabcentral/answers/132022-very-low-score-similarity-in-pca

Editada: primrose khaleed el 2 de Jun. de 2014

 hi everybody...i used pca to get similarity between images ..the problem is the score is very
low when enter the images to compering with the images in files
this is code

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] = princomp(images');

% step 5: only retain the top 'num_eigenfaces' eigenvectors (i.e. the principal components) num_eigenfaces = 2; evectors = evectors(:, 1:num_eigenfaces);

% step 6: project the images into the subspace to generate the feature vectors features = evectors' * shifted_images;

%%%%%%%%%%%%%%%%praper input imag%%% img3 = rgb2gray(imread('777.jpg'));

background = imopen(img3,strel('disk',15)); img3 = imabsdiff(img3,background);%strel,Create morphological structuring element (STREL)

myfilter = fspecial('gaussian',[5 5], 0.5);
myfilteredimage = imfilter(img3, myfilter, 'replicate');
%nois=imnoise(myfilteredimage,'salt & pepper');
medf= medfilt2(myfilteredimage,[8 8]);

%%%%%%%%%%%%%%%%%

se = strel('disk',4); img4 = edge(medf,'canny',0.08); img4= bwareaopen(img4,30);

input_image = imclose(img4,se); %closeBW=imfill(closeBW,'holes'); input_image=imresize(input_image,[80, 70]);

%%%%%%%%%%%%%%%%%%classifcation

% calculate the similarityty 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));

my quetions : why the score is very very low..(score =.0032)??? >how to specify the num_eigenfaces..(principal components, which is set by the variable num_eigenfaces)??? >>when decrease the num_eignface the score increase (why)??

plz help me

2 comentarios
Mostrar NingunoOcultar Ninguno

Image Analyst el 2 de Jun. de 2014

Haven't I ever asked you to read this before? If not, here it is again http://www.mathworks.com/matlabcentral/answers/13205-tutorial-how-to-format-your-question-with-markup

primrose khaleed el 2 de Jun. de 2014

Editada: primrose khaleed el 2 de Jun. de 2014

i used this code to get the features and get similarty between input images with images that stores in files:

 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] = princomp(images');
% step 5: only retain the top 'num_eigenfaces' eigenvectors (i.e. the principal components)
num_eigenfaces = 70;
evectors = evectors(:, 1:num_eigenfaces);
    % step 6: project the images into the subspace to generate the feature vectors
features = evectors' * shifted_images;
    % calculate the similarityty 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));