Replace picture's each 2*2 pixel values into their average value

Hi, I have a problem. I want to go with a 2*2 pixels moving window on a picture and take a snapshot then replace these 2*2 pixels' values with their average value -> move to next 2*2 until the whole picture is replaced. Can you guys write a brief code for me? Thx!!

Answer 1

[Image Analyst]

blurredImage = conv2(double(grayImage), [.25,.25;.25,.25], 'same');

Answer 2

[Youssef Khmou]

hi Jackson, This technique however does not produce good filtering, but here is standard code :

=================================================================

function ZZ=averaging_2x2(I)

ZZ=zeros(size(I));

I=im2double(I);

M=size(I);

for x=1:M(1)-1

    for y=1:M(2)-1

        ZZ(x,y)=I(x,y)+I(x,y+1)+I(x+1,y)+I(x+1,y+1);
        ZZ(x,y)=ZZ(x,y)/4;

    end
end
========================================================================

Note that this code does not give the same result as that given by Image Analyst , lets take an example :

===============================================

I=im2double(imread('circuit.tif'));

I=imnoise(I,'Gaussian');

X1=averaging_2x2(I);

X2=conv2(double(I),[.5,.5;.5,.5], 'same');

subplot(121), imshow(X1), subplot(122), imshow(X2) =================================================

try to manipulate the two codes and conclude.

KH.Youssef

[Image Analyst]

I mistakenly had 0.5 as the kernel elements initially, instead of 1/4. If you change to 0.25 they give the same values. conv2() is a lot more efficient, more so as the window size gets larger. You could also use imfilter().