How do i adjust the spacing between the slices of my stacked images?

Hi, i want to adjust the spacing between the slices of images i load in with "cat". Right now, they are stacked from 1 to 4, but how do i achieve that they are stacked not in steps of 1 but e.g. of 4?

What i did so far is this:

IA = imread('/Users/Desktop/Testdateien/1.tif')

IB = imread('/Users/Desktop/Testdateien/2.tif')

IC = imread('/Users/Desktop/Testdateien/3.tif');

ID = imread('/Users/Desktop/Testdateien/4.tif')

rgb4D = cat(4, IA, IB, IC, ID);

[D, map] = gray2ind(rgb4D);

I think i have to do sth. with the D matrix, but i don't know what.

Can anybody please help? Thanks in advance.

Answer 1

[Walter Roberson]

gray2ind does not apply to stacked images. You would have to do that and then stack the results. You would need to pass the rgb or gray table to gray2ind so that you got the same meaning of ind in each case.

To stack with layers of zeros, construct a layer of zeros of the appropriate size and stack it in.

spacing = 4;
Z = zeros([size(IA), spacing-1], class(IA)) ;

cat(4, Z, IA, Z, IB, Z, IC, Z, ID)

[Mike Schneider]

Hi, thank you this works pretty well, but i think i wrote the question wrong.

I only have meant to change the "labeling" of the z axis, so that the space between the layers stays the same, but they are not listet from 1 to 4, but from e.g. 4 to 16 (4 steps for each slice), because i want the spacing to keep the same.

So i found out that i can do this with the help of "axis", but i don't know how in detail. I am sorry but hope that you can help me anyway, thank you very much!

[Walter Roberson]

You have two questions here. One is about converting those images to indexed images in a consistent way. The other is about displaying images in some undefined way so that the Z coordinates show up differently.

To answer about displaying the Z coordinates, we need to know how you are displaying the stacked images.

I am not sure why you want to convert your images to indexed images. Indexed images cannot easily be processed mathematically because adjacent indices like 17 and 18 might refer to entirely different color or brightnesses. Indexed images can be recolored easily by changing the active colormap. However, "grayscale" images can also be easily recolored by changing the active colormap, so it is usually better to work with grayscale than with indexed.

gray2ind() is not able to process multiple images at the same time. gray2ind() also has no way of passing in a colormap to match against, so two different grayscale images could get converted to with different indexing schemes.

We cannot tell whether your tif images are grayscale or RGB.

num_index_colors = 64;   %how many indexed colors do you want?

IA = imread('/Users/Desktop/Testdateien/1.tif');
IB = imread('/Users/Desktop/Testdateien/2.tif');
IC = imread('/Users/Desktop/Testdateien/3.tif');
ID = imread('/Users/Desktop/Testdateien/4.tif');

imgs = {IA, IB, IC, ID};
nimgs = length(imgs);
ind_imgs = cell(1,nimgs);

cmap = gray(num_index_colors);

for K = 1 : nimgs
  this_image = imgs{K};
  if ndims(this_image) == 1
    this_image = this_image(:,:,[1 1 1]);  %rgb equivalent
  end
  ind_imgs{K} = rgb2ind( this_image, cmap );
end

image_stack = cat(4, ind_imgs{:});

[Mike Schneider]

Hi, i am very sorry because i am new to Matlab and therefore have very low experience in the image processing.

I am working with grayscale images and i am trying to do sth. like in the "mri example" in Matlab, where i try to visualize a 3d volume with a set of 2d slices(images) taken from the volume with a distance in the z axis of about 150 micrometer. Therefore i am trying to visualize sth. later with "isosurface" to "interpolate" the missing points in between the slices.

In the x and y direction Matlab uses the pixels for scaling the axes, but e.g. i know that 120 pixels are 500 micrometer, so how can i convert this. I have the same question for the z direction, where the images i load in with "cat" are only stacked from 1 to x, depending on how much images i stack. Here i would like to have the images stacked also in distances of e.g. 150 micrometers but with the same ratio as in the x and y axis, so that they are plotted in the right ratio for a later volume calculation or similar.

[Walter Roberson]

If you are looking at the example https://www.mathworks.com/help/images/examples/exploring-slices-from-a-3-dimensional-mri-data-set.html then in step 3 where they talk about

"We create a tform via composition starting with a 2-D affine transformation T1 that scales the (new) dimension 1 by a factor of -2.5 and adds a shift of 68.5 to keep the array coordinates positive"

then the -2.5 corresponds to the pixel spacing. So you would alter

T1 = maketform('affine',[-2.5 0; 0 1; 68.5 0]);

to reflect the spacing you want, and then you would put the result through the resampler.

Answer 2

[Massimo Zanetti]

First, most probably you don't want to stack them in the 4-th dimension as you wrote

rgb4D = cat(4, IA, IB, IC, ID);

Because this is not RGB image. If I have to guess what you are trying to do, I come with this:

IA = imread('/Users/Desktop/Testdateien/1.tif')
IB = imread('/Users/Desktop/Testdateien/2.tif')
IC = imread('/Users/Desktop/Testdateien/3.tif');
ID = imread('/Users/Desktop/Testdateien/4.tif')
rgb4D = cat(3, IA, IB, IC, ID); %notice the 3 instead of 4
%if IA,...,ID are single band images, to stack them with spacing of 4
STACKED4 = zeros([size(IA),13]);
STACKED4(:,:,1:4:13) = rgb4D