Problem with MATLAB functions - function

Question

I got this MATLAB function. When I ran it, the following error messege showed up. Can anybody give me some hint? Thank you. The code is also shown below.
[h,im_matched,theta,I,J]=im_reg_MI('keyframe1.jpg','keyframe2.jpg', 0, 1)
??? Undefined function or variable "h".
Error in ==> im_reg_MI at 74
[a, b] = max(h(:));% finding the max of MI and indecises
Below is the code.
[h,im_matched, theta,I,J]=im_reg_MI(image1, image2, angle, step)
[m,n]=size(image1);
[p,q]=size(image2);
[a,b]=size(angle);
im1=round(image1);
for k=1:b
J = rotate_image(angle(k),image2); %rotated cropped IMAGE2
image21=round(J);
[m1,n1]=size(image21);
for i=1:step:(m1-m)
for j=1:step:(n1-n)
im2=image21(i:(i+m-1),j:(j+n-1)); % selecting part of IMAGE2 matching the size of IMAHE1
im2=round(im2);
h(k,i,j)=MI2(im1,im2); % calculating MI
end
end
end
[a, b] = max(h(:));% finding the max of MI and indecises

Answer 1

The problem is that you aren't actually passing in image data for the image1 and image2 arguments. You're just passing character strings containing the image file names 'keyframe1.jpg' and 'keyframe2.jpg'.
You need to load the image data from the files first using IMREAD, then pass the image data to im_reg_MI. Assuming the images are in the current working directory, you would do something like this:
image1 = imread('keyframe1.jpg');
image2 = imread('keyframe2.jpg');
[h,im_matched,theta,I,J] = im_reg_MI(image1,image2,0,1);
EDIT:
There seems to be an additional error within im_reg_MI that occurs if image21 (the rotated version of image2) is the same size or smaller than image1 for one or more of its dimensions. If m1 were less than or equal to m and/or n1 were less than or equal to n, then one or both of the loops for i or j would never be entered since 1:step:(m1-m) and/or 1:step:(n1-n) would create an empty vector. Thus, the inner loop code would never be run and h would never be created.
And one extra note...
I noticed that the function im_reg_MI appears to treat the two images as 2-D, which means they must be intensity images (i.e. grayscale or binary image data). If you're dealing with indexed or RGB image data, I don't think im_reg_MI is going to handle them properly.