how to differentiate low order and high order fft coefficients of an image

I am working on fft compression so the fft signals are passed through lpf so that I can zerodown high frequency signals.But I am not able to differentiate Low and high frequency fft signals from lpf output.How to do that?Help me out

[Anitha Raj]

%This code is used to Butterworth lowpass filter close all; clear all; clc; I=imread('Sunflower.bmp'); R = I(:,:,1); G = I(:,:,2); B = I(:,:,3);

fc=50;%Cutoff frequency n=0.9; [co,ro,so] = size(I); cx = round(co/2); % find the center of the image cy = round (ro/2); cz = round(so/2); imf1=fftshift(fft2®); imf2=fftshift(fft2(G)); imf3=fftshift(fft2(B)); H=zeros(co,ro); for i = 1 : co for j =1 : ro d = (i-cx).^2 + (j-cy).^ 2 ; H(i,j) = 1/(1+((d/fc/fc).^(2*n))); end; end; H1=double(H<=fc); outf1 = imf1.*H1; % Fs=100; %sampling frequency % T=1; %signal length % N=T*Fs; %number of samples % f=-Fs/2:Fs/N:Fs/2-Fs/N; %frequency vector % plot(f,abs(outf1)) % outf1(n/2+1-20:n/2+20) = zeros(40,1) % Maxium1=max(max(max(abs(outf1)))); % Minium1=min(min(min(abs(outf1)))); % Th1=0.01*Maxium1;

%to set below threshold values to zero % outf1=outf1.*(outf1>Th1);

 % To filter zeros from matrix
 pos1=find(outf1<=2000);
pos1(pos1==0)=[];% To remove zeros

outf2 = imf2 .* H1; pos2=find(outf2<=2000); pos2(pos2==0)=[] outf3 = imf3 .* H1; pos3=find(outf3<=2000); pos3(pos3==0)=[] Nz=length(pos1+pos2+pos3); CR=(m*n*3)/size(Nz);% compression in transform domain defined as ratio of image size to number %of non zero fft co efficients out1 = abs(ifft2(outf1)); out2 = abs(ifft2(outf2)); out3 = abs(ifft2(outf3)); out=cat(3,out1,out2,out3); imshow(I),title('Original Image'),figure,imshow(uint8(out)),title('Lowpass Filterd Image');

This is my code.Input is color image and fft is found , then passed through lpf.The output of lpf contains both low frequency & high frequency contents so I am zerdowning high frequency to perform compression.Please see where Iam going wrong? 0 Comments Anitha Raj

Answer 1

[Image Analyst]

They occur at different indexes. Just display the fft image and you'll see.

[Anitha Raj]

Thank you for the last answer.I want to zero down high frequency signals and retain low frequency signals so that I can reconstruct signal using IIFT.But how to do that?Please help me I am struck at this stage.

[Image Analyst]

See my two attached demos.