MATLAB: Converting JPG images into a cell arrays

cell arraysdeep learningimage processingmachine learning

Hi
I have the following code i want to convert images jpg (TrainingData) into cell arrays ,where each cell containing a 28-by-28 matrix representing a synthetic image of fingerprint 
dataDir= fullfile('Data/');
exts = {'.jpg','.png','.tif','BMP'};
 
 
imds = imageDatastore(fullfile(dataDir),...
'IncludeSubfolders',true,'FileExtensions','.jpg','LabelSource','foldernames');
countEachLabel(imds);
 
[TrainData, TestData] = splitEachLabel(imds, 0.5);
size(TrainData);
countEachLabel(TrainData);
for i = 1:numImages
    img = readimage(TrainData, i); 
    
    img  = imbinarize(img); 
      
    
end

Best Answer

sz = size(img);
RB = 28; CB = 28;
NRB = floor(sz(1)/RB);
LOR = sz(1) - NRB*RB;
NCB = floor(sz(2)/CB);
LOC = sz(2) - NCB*CB;
if LOR ~= 0
  rbs = [RB * ones(1,NRB), LOR];
else
    rbs = RB * ones(1,NRC);
end
if LOC ~= 0
    cbs = [CB * ones(1,NCB), LOC];
else
    cbs = CB * ones(1,NCB);
end
tiles = mat2cell(img, rbs, cbs, size(img,3));
See also mat2tiles in the File Exchange
And watch out for the titles that are not 28 x 28. This code does not assume that the image is an exact integer multiple of 28 in each direction, and it does not throw away any partial blocks.
Related Question