Save one image as part of a larger image

Hi all,

I have processed image data into a 10x512 pixel array. I have multiple images all this size and I would like to display them all side by side on a single larger image. I have created a blank image using zeros of size 1000x512 (i.e. I have 100 images in total, therefore 10 pixels x 100 images) but am having trouble getting my smaller image to display. How could I approach this?

Kind regard and thanks, Susan

Answer 1

[Geoff]

Let's just assume that all your images are in a 3D matrix and you want to slap them onto a canvas:

canvas = zeros(1000, 512);    % Target image
images = rand(10, 512, 100);  % 100 very useless images =)

Now, in this case, there are clever MatLab functions, but I'll stick to a simple loop so you see what's going on (and that's adaptable if your images are stored in some other way).

for n = 1:100
  rows = (n-1) * 10 + (1:10);
  canvas(rows, :) = images(:,:,n);
end

So, for each image (indexed by n), you calculate a range of row indices for your 10 rows. The value (n-1) * 10 gives the zero-based position of the first row of the 'n'th image, and then you add that to the vector 1:10. ie image 1 is rows 1:10, image 2 is rows 11:20, etc...

The slice canvas(rows, :) is then a 10-by-512 address, into which you assign the values from your 'n'th image (also 10-by-512).

[Susan]

Hi Geoff,

Thanks for your reply - I adopted your method and it works!! :)

However, I've still a problem. I'm processing 500 images in total - 10 (i_B in code) groups of 50 (num_A in code) - producing an image for each iteration of i_B (yielding 10 images in total of pixel size (N,50)). The code works fine if I set i_B = 1, 2, 3, etc. manually but I would like to automate it (so that the code will output and save 10 images in this case) as I will have a larger number of images in the future. The error I am getting is:

??? Subscripted assignment dimension mismatch

at the portion of the code below indicated by ******.

I would greatly appreciate any advice!

Kind regards,

Susan

P.s. I've omitted some of the preceding code for clarity (will not affect operation)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

num_B = 10

num_A = 50;

for i_B = 1:num_B

for i = 1:num_A

file = sprintf('%s%s%d%s',folder,filename,((100*(i_B-1))+1+(2*(i-1))),suffix);

A = imread(file);

mA = mean(A);

mA(1) = mA(2);

fit = polyfit(nm,mDC,4);

pDC = polyval(fit,nm);

S = mA./pDC;

S = S - mean(S);

Ses = interp1(nu,S,kes);

abs_s_tau = abs(fft(Ses));

img(:,i) = abs_s_tau'; ******

end

img = img(1:(pixel/2),:);

img = log10(img);

img = mat2gray(img);

img = img(1:pixel/2,:);

figure(i_B); clf

imshow(img);

file_saved = strcat(folder,'B-scan_',num2str(i_Bscan),'.png');

imwrite(img, file_saved ,'png', 'BitDepth', 16);

end

Answer 2

[Susan]

Just cracked it myself. Simple mistake. Needed to initialise the image inside the first for loop. :)

num_B = 10 
num_A = 50;

    for i_B = 1:num_B

    img = zeros(pixel,num_A);
    img(i_B);

        for i = 1:num_Ascans
  .
  .
  .
  .
        end
  end