How to reduce processing time for imread and imwrite

Question

I am working on a 3d reconstruction using parallel computing in Matlab. I profiled my code and found that imread and imwrite took most of the processing time. My goal is to significantly reduce the processing time as my simulations involve large data sets and iterations. My code:

projection_length = 4100;
parfor q = 1:projection_length
    tmpData = zeros(1600, 500);
    for i = 1:500
        fname= sprintf('pre%03d.tif', i);
        tmpData(:, i) = imread(fname, 'PixelRegion', {[1 1600], [q q]});

        disp(['Analyzing projection ' num2str(q)  ' of ' num2str(projection_length) ', Angle ' num2str(i) '...']);
    end
    idata=255-tmpData;
    H = iradon(idata, 0.72, 'Hann', 0.8, 1600 );
    postfname= sprintf('post%06d.tif', q);
    imwrite(H, postfname);
end

Answer 1

It looks like you read the same 500 images over and over again, only to access different pixel regions. Better to read the images once, and then access the pixel regions using standard array slicing.

I can't really test the code because I don't have your image files and functions, but it should look something like this:

% preload images
images = {};
for i = 1:500
    fname= sprintf('pre%03d.tif', i);
    images{i} = imread(fname);
end
% analyze images
projection_length = 4100;
parfor q = 1:projection_length
    tmpData = zeros(1600, 500);
    for i = 1:500
        tmpData(:, i) = images{i}(1 : 1600, q, :);
        disp(['Analyzing projection ' num2str(q)  ' of ' num2str(projection_length) ', Angle ' num2str(i) '...']);
    end
    idata=255-tmpData;
    H = iradon(idata, 0.72, 'Hann', 0.8, 1600 );
    postfname= sprintf('post%06d.tif', q);
    imwrite(H, postfname);
end