How to optimize for loop

Hi all..i have a 1424x2144 image.i want to process each pixel in the image to find a new pixel value.And i want to create a new 1424x2144 matrix for the new image. I have the following code

if true
  for i=1:1424
     for j=2:2144
         finding new pixel values.
        newimg(i,j)=value;
     end
  end    
end

But it takes about 30 seconds to complete the iterations. How to speedup the execution time.?

[Adam]

Clearly it depends what 'finding new pixel values' does/means.

As it is you seem you be just assigning the same value to every pixel which you can do in 1 line very quickly.

Answer 1

[Jos (10584)]

I think I am missing something. Why runs j from 2 instead of 1? Is value a constant? If so, this would suffice:

newimg = repmat(value,1424,2144) ;
newimg(:,1) = 0 ;

In any case, if you use for-loops to create a new matrix, you can speed things up tremendously by pre-allocating the matrix. In your situation:

 newimg = zeros(1424,2144) ; % pre-allocation with zeros
  for i=1:1424
     for j=2:2144
         finding new pixel values.
        newimg(i,j)=value;
     end
  end

If you have looked carefully, the matlab editor warns for this. There is a red line under newimg and when you hover over it with your cursor it a message pops up " The variable appears to be growing inside a loop ...".It is gives a suggestion to fix it :)