How do I make the reconstructed light image more clear with less noise?

2 vues (au cours des 30 derniers jours)
Diptayan Dasgupta
Diptayan Dasgupta le 13 Mai 2021
Source image
I just want to make the reconstructed light image more clear and more visible if its possible. That's the code I used. The above image is the source.
clear all, close all;
Ii=imread('1 A.png');% 256*256 pixels 8bit
figure; imshow(Ii);
title('Object pattern')
axis off
Ii=double(Ii);
Ii=Ii(:,:,1);
PH=rand([256,256]);
Ii=Ii.*exp(2i*pi*PH); % add a random phase on the object
M=512;
I=zeros(512);
I(128:383,128:383)=Ii; % zero padding
z=1; %(cm, distance) PLEASE CHANGE THE DISTANCE TO 1, 5, 15, ETC.
w=6500*10^-8; %(cm, wavelength)
delta=0.005; % cm, pixel size 50um
r=1:M;
c=1:M;
[C, R]=meshgrid(c, r);
% Forward propagation (650nm)
p=exp(-2i*pi*z.*((1/w)^2-(1/M/delta)^2.*(C-M/2-1).^2-(1/M/delta)^2.*(R-M/2-1).^2).^0.5);
A0=fftshift(ifft2(fftshift(I)));
Az=A0.*p;
E=fftshift(fft2(fftshift(Az))); % 1st order of the hologram
% Reconstruction (650nm)
p=exp(-2i*pi*(-z).*((1/w)^2-(1/M/delta)^2.*(C-M/2-1).^2-(1/M/delta)^2.*(R-M/2-1).^2).^0.5);
A1=fftshift(ifft2(fftshift(E)));
Az1=A1.*p;
R1=fftshift(fft2(fftshift(Az1)));
R1=(abs(R1)).^2;
figure; imshow(R1/max(max(R1)));
title('Reconstructed image(650nm)')
axis off
% Reconstruction (450nm~650nm)
dw=50;
IMA=zeros(512,512);
for g=0:40;
w2=(20000-dw*g)*10^-8; % reconstruction wavelength
E2=E.*exp(2i*pi*sind(10)*(w-w2)/w/w2.*R*delta);
% phase mismatch due to the wavelength shift
p=exp(-2i*pi*(-z).*((1/w2)^2-(1/M/delta)^2.*(C-M/2-1).^2-(1/M/delta)^2.*(R-M/2-1).^2).^0.5);
Az2=ifft2(fftshift(E2)).*(fftshift(p));
R2=fftshift(fft2(Az2));
R=(abs(R2)).^2; % summation of all wavelengths
IMA=IMA+R2;
end
IMA=IMA/max(max(IMA));
figure; imshow(IMA)
title('Reconstructed image(white light)')
axis off
Reconstructed white light image
  2 commentaires
DGM
DGM le 13 Mai 2021
You might want to attach the source image so that others can test your code with it.
Diptayan Dasgupta
Diptayan Dasgupta le 13 Mai 2021
I just attached the source image @DGM

Connectez-vous pour commenter.

Réponses (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by