Matlab alpha trimmed filter

I am implementing this but i am getting the result full blank help me with this I have to create the filter manually and apply using for loop I am only allowed to use sort function I am not allowed to use the build in functions I have attached the picture please help me with this Pic :

[Geoff Hayes]

Rashid - please post the code that you have written so that we can better understand where the problem exists.

[Rashid Hussain]

d1=8;
d2=8;
n=5;
I = imread('pout.tif');
I = double(I(:,:,1))/255;

% images, 1. whitenoise, 2. salt&pepper, 3. white and s&p.
img = imnoise(I,'gaussian',0,0.005);

if (isa(img,'uint8'))
  % force img to be bouble between 0.0 .. 1.0
 img=double(img)/255;
else
  % assume img is double in range 0..1 
end

% create a blank array for output
img_alpha = zeros(size(img));

% --- YOUR CODE TO GO HERE ---

% NOTE: your solution does NOT have to deal with values near the image boundary

for r=1:size(img,1)
  for c=1:size(img,2)

      if r-n < 1 | n+r >= size(I,1) | c-n < 1 | c+n > size(I,2)
      img_alpha(r,c)=img(r,c);
      else
          fromrow=abs(r-n);
          torow=r+n;
          fromcol=abs(c-n);
          tocol=c+n;

           neig=img(fromrow:torow,fromcol:tocol);
      localval=neig(:);
      s=sort(localval);
      sumofval=0;
      for i=d1+1:size(s,1)-d2

          sumofval=sumofval+s(i);
      end

      avg=(1/(n*n)-(d1+d2))*sumval;
      img_alpha(r,c)=avg;

      end

    end
  end

imshow(img_alpha);

[Rashid Hussain]

Please tell me whether I am doing right or wrong.

I am not allowed to use built-in functions, and you can see the pictures for problem description. Thanks

[Image Analyst]

http://www.mathworks.com/matlabcentral/answers/38787-what-can-be-programmed-without-any-built-in-functions

Answer 1

[Image Analyst]

sumval is a different variable than sumofval. Try using sumofval instead, and then the code will not throw an error.