Standard Deviation Between images

Hi there,I currently have set of images which I get from taking the average from a group of smaller images, I am just wondering if it is possible if I could get an array of the std deviation between the images from n2 to n1? Everytime I try I end up taking just the Overall Standard Deviation from the image

if true
   for n=n1:n2
  stem='E:\29th April New\NEW Gain=99_No Delay_ Gate=0.02_0';
  r=n-1;
  r=int2str(r);
  combinedStr=strcat(stem,r);
  Images=99;
  t=10;
AvgImg =(zeros(1024,1280)); 

    for i=t:Images;
        B=readimx(fullfile(combinedStr,['B000',int2str(i),'.im7']));
        C=B.Frames{1}.Components{1};
        V = C.Planes;
        Img = V{1,1};
        J = imrotate(Img,-90);
        I2 =(flip(J,2));
        FUNC=@(x)max(x(:));
        I3=medfilt2(I2,[1 1]);
        AvgImg = AvgImg +double(I3);
    end
    figure(n-1)
    AverageImg = AvgImg/(Images-t);
   AverageImgB=90;
    AverageImgLII=uint16(AverageImg-AverageImgB);
    K=imagesc(flipud(AverageImgLII));
    set(gca,'YDir','normal');
  end

Answer 1

[Guillaume]

Store your images in a 3D matrix and calculate all your stats in one go using mean, std, etc. on the 3rd dimensions:

numimages = 99
stem = 'E:\29th April New\NEW Gain=99_No Delay_ Gate=0.02_0';
for gating = n1:n2
  stemgating = sprintf('%s%d', stem, gating-1);
  LIIimages = cell(1, numimages);  %storing in a cell array, to be converted to 3D matrix later on
  for imgindex = 1:numimages
     imxstruct = readimx(fullfile(stemgating, sprintf('B%05d.im7', imgindex)));
     img = imxstruct.Frames{1}.Componenets{1}.Planes{1};
     LIIimages{imgindex}= imrotate(img, -90);
  end
  %convert cell array to 3d matrix
  LIIimages = cat(3, LIIimages{:});
  %calculate stats:
  meanimg = mean(LIIimages, 3);
  stdimg = std(LIIimages, 0, 3);
  %... do more stuff
end

Note that I'm storing the images temporarily into a cell array. This avoids having to hardcode the size of the images (the alternative is to predeclare the 3D array with zeros(1024, 1280, numimages))

Also note, that I've fixed the problem with you not being able to create properly the name of the first 9 buffer images by using sprintf with a format string that says to prepend the number with enough zeros to have 5 digit (the '%05d')

[Robert Roy]

Hi there, that seems brillant thanks very much. Only problem I am having now is that it is saying Reference to non-existent field 'Componenets'.

Error in Experimental (line 27) img = imxstruct.Frames{1}.Componenets{1}.Planes{1};

I am unsure what that means apologies

[Guillaume]

That means I misspelled Components. It's the same code that you wrote to load the Davis image, only streamlined.

Also, note that I omitted the median filter, since a median filter with a 1x1 kernel does not do anything.

[Robert Roy]

Ah sorry I didnt see that. Apologies I am also now getting a message saying The right hand side of this assignment has too few values to satisfy the left hand side.

Error in Experimental (line 28) LIIimgages{img}=imrotate(img, -90); I dont see where that is coming from

[Guillaume]

That error is because I mistakenly used the same variable name for two different things. On top of that I misspelled LIIimages.

I've fixed my original answer. Hopefully, there's no more typos.

[Robert Roy]

Thats brillant thanks very much, one of the things I would also like to do is to plot each individual image gating to compare against the average value, is there a way to do this?

[Guillaume]

I'm not sure what you mean by plotting an image against another.

Note that all the images are held in LIIimages. image number n is LIIimages(:, :, n). You can display these however you want.