Variable Number of images as inputs and superimpose together? - Homework

I am trying to make a function that will receive a variable number of image files. The function will read all the images and superimpose all images with equal contribution from each. So the input images might be different sizes, so the function needs to resize all images to the same size as the smallest image before superimposing.

This is the function I have so far, but it needs a lot more work.

function SuperImpose( varargin )
% INPUT VARIABLE NUMBER OF IMAGES AND USE imlincomb TO COMBINE THEM
%   17-NOV-2013, NM

cn = 0;

for i = 1:nargin
cn = imresize(varargin{i}, [384 512]);
cn = cn + 1;
cs{cn} = str;

end
sumim = imlincomb(cn);
imshow(sumim);

end

Answer 1

[Image Analyst]

You need two loops. One to read in all the images and see what sizes they are and to store the size of the smallest. Then another one where you read them all in and resize them according to the smallest, and sum them. Something like this (untested):

smallestRows = inf;
smallestColumns = inf;
for i = 1:nargin
  thisImage = imread(varargin{i});
  [rows, columns, numberOfColorChannels] = size(thisImage);
  if rows < smallestRows
    smallestRows = rows;
  end
  if columns< smallestColumns
    smallestColumns= columns;
  end
end
theSum = zeros(smallestRows, smallestColumns, numberOfColorChannels);
for i = 1:nargin
  thisImage = imread(varargin{i});
  theSum = theSum + double(imresize(thisImage, [smallestRows, smallestColumns]));
end
imshow(theSum, []);

To make it more robust, check to make sure that all images have the same number of color channels, and if they don't convert the gray scale images to color before summing.

[Nora]

What does the

theSum = zeros(smallestRows, smallestColumns, numberOfColorChannels);

do? is 'numberOfColorChannels' a function?

[Image Analyst]

It initializes a sum array to the size of the smallest image you encountered during the first loop, so that you can sum the resized images into it. Look where numberOfColorChannels occurs - it's the output of the size() function. You can tell by it's name that it is the number of color channels the image has, right? So it's 1 for a grayscale image and 3 for an RGB image.

[Nora]

I see! The reason I ask is because I got this error:

Where is this coming from?

EDU>> SuperImpose('lighthouse.jpg','Tulips.jpg','Desert.jpg')
Undefined function 'resize' for input arguments of type 'uint8'.
Error in SuperImpose (line 33)
  theSum = theSum + double(resize(thisImage, [smallestRows, smallestColumns]));

[Nora]

I changed resize to imresize. Now I get a single blank image. Can you think of anything that would cause this?

Answer 2

[Walter Roberson]

I would suggest that instead of using varargin, that you pass a cell array containing the data.

Are you being passed the matrices of image data, or the file names ?

[Nora]

An example of calling the function is this:

>> im=SuperImpose('lighthouse.jpg','Tulips.jpg','Desert.jpg');  
>>imshow(im);