How can you take a maximum intensity projection (MIP) of an image stack?

So I'm pulling an image stack from a filepath (all jpegs) and combining them to form a an image stack. I then process the stack through the a grayscale protocol (basically rgb2gray for multiple images at the same time). I then run the graystack through the max command down each column along the Z-axis. I convert the result X x Y matrix into a 8 bit image with uint8 command. However, my resulting image is not an MIP. It's overly saturated (filled with white pixels), especially in comparison to the MIP I acquire when I run the stack through ImageJ. I believe my computational method for MIP is incorrect and may very well over-simplify what an MIP mathematically does. I was wondering if you had any suggestions? Many many thanks!

 for k=1:9
    jpgFileName=strcat('Image0090_01000',num2str(k),'.jpg');
    if exist(jpgFileName,'file')
        imageData=imread(jpgFileName);
        for i=1:b
            for j=1:b
                Stack(i,j,1,k)=imageData(i,j,1);
                Stack(i,j,2,k)=imageData(i,j,2); %populate Stack with RGB intensity values 
                Stack(i,j,3,k)=imageData(i,j,3);
            end
        end

    else
        fprintf('File %s does not exist.\n',jpgFileName);
    end
end
for k=10:99

    jpgFileName=strcat('Image0090_0100',num2str(k),'.jpg');
    if exist(jpgFileName, 'file')
        imageData=imread(jpgFileName);

         for i=1:b
            for j=1:b
                Stack(i,j,1,k)=imageData(i,j,1);
                Stack(i,j,2,k)=imageData(i,j,2);
                Stack(i,j,3,k)=imageData(i,j,3);
            end
        end

    else 
        fprintf('File %s does not exist.\n',jpgFileName);
    end 
end
for k=100:a
     jpgFileName=strcat('Image0090_010',num2str(k),'.jpg');
    if exist(jpgFileName, 'file')
        imageData=imread(jpgFileName);

         for i=1:b
            for j=1:b
                Stack(i,j,1,k)=imageData(i,j,1);
                Stack(i,j,2,k)=imageData(i,j,2);
                Stack(i,j,3,k)=imageData(i,j,3);
            end
        end

    else 
        fprintf('File %s does not exist.\n',jpgFileName);
    end 
end

grayStack = squeeze( 0.2989 * Stack(:,:,1,:) + 0.5870 * Stack(:,:,2,:) + 0.1140 * Stack(:,:,3,:) ); %X, Y, Z after squeeze--this is the RGB2gray test=int8(max(grayStack,[],3)); %%I feel like this bit conversion might be where my calculations be incorrect imshow(test);

Answer 1

[Image Analyst]

You don't need loops over i and j, just operate on the whole 3D image with the max() function. Also, you need only one loop over k if you just get the filenames before the k loop with the dir command. After you get the max RGB image, you can convert to grayscale if you want to.

[Harsh Patolia]

Thank you so much for your help. The only way I can construct the image stack, however, is with my for loops with i and j. MATLAB does not import images directly as image stacks does it? I was under the impression it only does so as sequences, and the user is required to program a code to construct an image stack from these sequences? Thank you, once again!

[Image Analyst]

It can create a stack from a sequence of files. And, like I said you don't need loops over i and j, you only need k. Also, do you want your max to be color or gray scale?