MATLAB: Subtracting mean image intensity value from all individual pixels

image analysisimage processingImage Processing Toolbox

I have 3D imaging stacks that I'd like to investigate individual images from within the image stacks. I have found the mean intensity for each image and now would like to subtract this mean from all pixels within that image. I want to do this for all images within the stack and then show the image stack. I've tried using imsubtract, but I'm getting matrix mismatch errors. Here's the code in question.
% Request file selection.
[fileNameF pathNameF] = uigetfile('*C0*.tiff');
filePathF = fullfile(pathNameF, fileNameF);
disp(fileNameF);
filePathB = strrep(filePathF, 'C0', 'C1');
%Input wavelength, etc.
prompt = {'Calibration ratio:', 'Lower threshold', 'Upper threshold', 'Z step size [um]'};
dlg_title = 'Inputs';
num_lines = 1;
if exist('fbdefaults2.mat', 'file')
    load('fbdefaults2.mat', 'def');
else
    def = {'1','6', '254', '1'};
end
answer = inputdlg(prompt,dlg_title,num_lines,def);
fbCalibrationRatio=str2num(answer{1});
lowerThreshold=str2num(answer{2});
upperThreshold=str2num(answer{3});
zStepSize=str2num(answer{4});
def=answer;
save('fbdefaults2.mat', 'def');
info=imfinfo(filePathF);
slices = length(info);
height=info(1).Height;
width=info(1).Width;
fAttVecMean=zeros(1,slices);
bAttVecMean=zeros(1,slices);
newIfmean=zeros(1,slices);
for slice=1:slices
    If = (imread(filePathF, slice))
    Ib = (imread(filePathB, slice))
      %Thresholding the images within the stack
      mask = find((If < lowerThreshold) | (If > upperThreshold));
      If(mask)=NaN;
      %Ib(mask)=NaN;%Stringent condition.
      mask = find((Ib < lowerThreshold) | (Ib > upperThreshold));
      Ib(mask)=NaN;
      %If(mask)=NaN;%Stringent condition.
      fAttVal = nanmean(If(:));
      bAttVal = nanmean(Ib(:));
      fAttVecMean(1,slice)=fAttVal;
      bAttVecMean(1,slice)=bAttVal;
      IfDif(1,slice)=imsubtract(If,fAttVecMean);
  end
Thanks for the assistance!

Best Answer

There is absolutely no need to mess around with nans, find(), and nanmean(). Just simply do this:
[rows, columns, numSlices] = size(If);
IfDif = zeros(rows, columns, slices, class(If));
for slice=1:slices
    If = (imread(filePathF, slice))
    mask = If < lowerThreshold) | (If > upperThreshold);
    fAttVecMean(slice) = mean(If(mask));
    Ib = (imread(filePathB, slice))
    mask = Ib < lowerThreshold) | (Ib > upperThreshold);
    bAttVecMean(slice)= mean(Ib(mask));
    % Insert the difference image into the current slice of the 3D image.
    IfDif(:,:,slice)= If - fAttVecMean(slice);
end
And like I said earlier, call the image something, anything other than If - calling it "thisSlice" would be a conventional name that you could use.
Related Question