How to implementation DFT (discrete fourier transform) to intensity?

Here I am trying to calculate intensity of Xpolarized light by using DFT (discrete fourier

transform) instead of fft, by using DFT I am getting intensity in 4th quadrant only (Which is

shown in figure (1)). How can I get intensity in total plane. Can some help me in this regard

clear all;

clc;

lose all;

[kx,ky]=meshgrid(-1:2/21:1);

circ=sqrt(kx.^2+ky.^2)<1;

alp=asin(0.95);

k0=1/sin(alp);

kz=sqrt(k0^2-(kx.^2+ky.^2));

GX=sqrt(k0./kz).*((k0*ky.^2+kz.*kx.^2)./(k0*(kx.^2+ky.^2)));

GY=sqrt(k0./kz).*((kz-k0).*kx.*ky)./(k0*(kx.^2+ky.^2));

la=1;

lb=0;

EX=(GX.*la-rot90(GY.*lb, 3));

M=22;

N=22;

F=zeros(22,22);

for Xpx=1:M

for Xpy=1:N

F(Xpx,Xpy)=sum(sum(EX.*exp(1j*2*pi*(((Xpx.*kx)/M)+((Xpy.*ky)/N)))));

end

end

F=F/(M*N);

I1=F.*conj(F);

figure(1)

imagesc(I1),colormap gray;axis image; axis off;

Thank you very much in advance Your Mohd. Gaffrar

Answer 1

[Dishant Arora]

Use the logarithmic scale to image the fourier transform. Fourier transform results in large dynamic range, thereby logarithmic scale makes it easy to see small signal components in the presence of large ones. Probably you can do:

imagesc(log(I))

[MOHD]

Sorry

imagesc(log(I))

is not giving

[Dishant Arora]

The kx and ky you are using inside the loop for dft calualation, shouldn't they be integers. They need to be indices not spatial coordinates. Make appropriate changes to your code and cross check with built in fft2. I tried and it works for me.