MATLAB Code to read an image and subsample it to sizes 512, 256,128, 64,32.

clear

close all

Input = imread('Flower.jpg');

figure

subplot(2,3,1);

imshow(Input);

m = 1024;

s = num2str(m);

xlabel(s);

for i = 1:5,

    m = m/2;

    J=Input(1:2^i:size(Input,1),1:2^i:size(Input,2));

    subplot(2,3,i+1);

    imshow(J);

    s = num2str(m);

    xlabel(s); 

end;

gtext('using code');

% sampling using imresize()

m=1024;

scale=[512 256 128 64 32]

figure;

subplot(2,3,1);

imshow(Input);

s=num2str(m);

xlabel(s);

for i = 1:5

    m=m/2;

    subplot(2,3,i+1);

    imshow(imresize(Input,[scale(i) scale(i)]));

    s=num2str(m);

    xlabel(s);

end

gtext('using built-in function');

OUTPUT RESULT :

Sampling using Code

Sampling using built-in function imresize()We observe that sampling of an image with the use of manual code results in sharp edged images whereas the one with the built-in function produces a smoothed(blurry) images.

CONCLUSION:

Sampling process was carried out on an input image, subsampling its size to 512 × 512, 256 × 256, 128 × 128, 64 × 64 and 32 × 32 with the help of a built in function and a written code. The outputs are shown in Fig.1 and Fig.2. It can be seen that the output image remains nearly the same for 512 × 512, but a slow degradation starts for 128 × 128 and the image becomes virtually unusable for 32 × 32. Thus, it is safe to sample the image at half the original sampling rate as it does not cause any significant loss of information.