MATLAB: [HELP] Extract double array from cell

facelocal binary patternMATLABnxn regionregionsegmentation

I have an image which has been divided into 49 blocks. from the block I would like to get the feature vector of LBP from each cell. what can i do?
any advise is greatly appreciated.
 proceed = 1;
 data_matrix = [];
 ids = [];
 conut = 0;
 l=0;
 for i=1:100
        for j=1:10
 s = sprintf('database/s%i/%i.jpg',i,j);
           X = double(imread(s));
            [rows columns] = size(X);
            blockSizeR = 45; % Rows in block.
            blockSizeC = 34; % Columns in block.
          % Figure out the size of each block in rows. 
% Most will be blockSizeR but there may be a remainder amount of less than that.
            wholeBlockRows = floor(rows / blockSizeR);
            blockVectorR = [blockSizeR * ones(1, wholeBlockRows), rem(rows, blockSizeR)];
% Figure out the size of each block in columns. 
            wholeBlockCols = floor(columns / blockSizeC);
            blockVectorC = [blockSizeC * ones(1, wholeBlockCols), rem(columns, blockSizeC)];
% Dividing the image into 49 Blocks
                ca = mat2cell(X, blockVectorR, blockVectorC);
                eval(['out_' num2str(l) '=ca' ]);
                l=l+1;
              MAPPING=getmapping(8,'riu2');
             feature_vector = lbp(ca,1,8,MAPPING,'hist');
            feat_vect=feature_vector';
            data_matrix = [data_matrix,feat_vect];
        end 
end

Best Answer

I'm assuming that the lbp code you're referring to is from http://www.cse.oulu.fi/CMV/Downloads/LBPMatlab, which would indeed give you the error you mention if you pass it the cell array ca since it is expecting an image.
Possibly, what you want to do is:
feature_vectors = cellfun(@(cell) lbp(cell, 1, 8, MAPPING, 'hist'), ca, 'UniformOutput', false);
which will send each cell in turn to lbp and collate all the returned values in a cell array.
If you want to then stuff these return values in your data_matrix you'll have to change its declaration to:
data_matrix = {}; %cell array instead of matrix
and concatenate it with
data_matrix = [data_matrix {feature_vectors}];
or you could just do
data_matrix{l} = feature_vectors;
Related Question