Finding row-averaged magnitude spectrum

Hi, I have an image matrix of 512 x 512 and I'm trying to plot the row-averaged magnitude spectrum of this image. The following code is my first attempt to solve this problem. Can anybody help me whether I'm correct or not? Any help would be appreciated.

load blur.mat
    
    for i=1:512
        rows = blur(i,:)
        rowsFFT = fft(rows,512)
    end
    plot(abs(rowsFFT))

Answer 1

[Star Strider]

Assuming that I understand what you want to do, I would do something like this:

blur = rand(512);
L = size(blur,1);                                       % Row Size
rowsFFT = fft(blur)/L;                                  % Compute ‘fft’ Of Columns, Normalise By Row Size

Fs = 1;                                                 % Sampling Frequency (1/Pixel)
Fn = Fs/2;                                              % Nyquist Frequency
Fv = linspace(0, 1, fix(L/2)+1)*Fn;                     % Frequency Vector
Iv = 1:numel(Fv);                                       % Index Vector

figure
plot(Fv, abs(rowsFFT(Iv,:))*2)
grid
xlabel('Frequency (pixel^{-1})')
ylabel('Amplitude (Units)')

Alternatively:

rowsFFT = fft(blur-mean(blur));                         % Remove Constant Offset, Compute ‘fft’ Of Columns, Normalise By Row Size

If this is not what you want, please provide more information.

[YAGIZ AKALIN]

Sir, thank you so much for your reply. How can I plot the magnitude of the row-averaged spectrum in terms of frequency in pi units? You plotted in terms of pixel^{-1}.

[Star Strider]

My pleasure!

Replace ‘Fv’ with:

Fv = linspace(0, 1, fix(L/2)+1)*pi;                     % Frequency Vector

and the plot becomes:

figure
plot(Fv, abs(rowsFFT(Iv,:))*2)
grid
xt = 0:0.1:1;
xtl = compose('%.1f\\pi', xt);
xtl{1} = '0';
xtl{end} = '\pi';
set(gca, 'XTick',xt*pi, 'XTickLabel',xtl)
xlabel('Radian Frequency (rad)')
ylabel('Amplitude (Units)')
xlim([0 pi])

That should do what you want.

[YAGIZ AKALIN]

Thank you so much sir.

[Star Strider]

As always, my pleasure!