??? Subscript indices must either be real positive integers or logicals.

i am new and doing 2d convolution but i am stuck at this error. here is the code:

function output_args = convo(img,k )
%user giant malik
%UNTITLED4 Summary of this function goes here
%   Detailed explanation goes here
%I = [1 2 3;
%4 5 6;
%7 8 9;];
%k = [0 0 0;
%0 1 0;
%0 0 0];
[r1 c1]= size(img);
[r2 c2]=size(k);

 length = r1+(r2-1); % for creating new dimensiont
  row = length;       % 
col = length;
a= (r2-1)/2; b= (c2-1)/2; % start value for calculating convultion 
s= a*(-1)+1;
t=(b)*(-1)+1;
%disp(s);
output_args=zeros([row, col]);

 z=s;w=t;
    for x=0:length
        for y=0:length
           for s=s:a+1
              % disp(s);
               for t= t:b+1
                   output_args(x,y)=output_args(x,y) + (output_args(s, t)* f(z-s,w-t));   % error pointing here

               end
           end
     w=w+1;   
        end
        z=z+1;
    end

end

Answer 1

[Azzi Abdelmalek]

In your code there are indices equal to 0.

output_args(x,y)   % with x starting at 0 will give you an error, you should start at 1

[GIANT]

i have intialise all with 1 but now its giving "??? Undefined function or method 'f' for input arguments of type 'double'." can u tell me how shell i intialise "f"?

Answer 2

[Image Analyst]

Lots of problems with that code. No checking for number of dimensions of img. No initialization of output_args. No try/catch. Virtually no comments. No definition of the "f" function, Etc.

What happens if someone passes in a color image for img? Kaboom! Why? See this http://blogs.mathworks.com/steve/2011/03/22/too-much-information-about-the-size-function/

Put a try/catch into your function:

try
  % Some code that might generate an error.
catch ME
  errorMessage = sprintf('Error in function %s() at line %d.\n\nError Message:\n%s', ...
    ME.stack(1).name, ME.stack(1).line, ME.message);
  fprintf(1, '%s\n', errorMessage);
  uiwait(warndlg(errorMessage));
end

Move this to be first line, right after, or even right before, the try line:

output_args = [];

Now, .... the main reason for your error is that you used f, which is undefined . You're not really programming up a convolution correctly. Think about what indexes you should use. Perhaps it would be easier for you to extract a little sub-image from the main image, which is right under your sliding kernel window. Adding some comment before each line might help you figure out where you're going wrong. Give it a try and if you can't figure it out, let me know.

[Image Analyst]

By the way, why not simply use conv2() or imfilter()???