How to find 2d discrete wavelet transform(dwt) for true color image

hi friends, how to find 2d discrete wavelet transform for true color image in matlab.please give some code example

[hemalatha]

http://in.mathworks.com/matlabcentral/answers/30159-dwt2-on-an-rgb-image

refer this link..u can get some idea

[Fateme Jalali]

clc;
clear all;
close all;

%Read Input Image
Input_Image=imread('D:/mytest.png');

%Red Component of Colour Image
Red_Input_Image=Input_Image(:,:,1);
%Green Component of Colour Image
Green_Input_Image=Input_Image(:,:,2);
%Blue Component of Colour Image
Blue_Input_Image=Input_Image(:,:,3);

%Apply Two Dimensional Discrete Wavelet Transform
[LLr,LHr,HLr,HHr]=dwt2(Red_Input_Image,'haar');
[LLg,LHg,HLg,HHg]=dwt2(Green_Input_Image,'haar');
[LLb,LHb,HLb,HHb]=dwt2(Blue_Input_Image,'haar');

First_Level_Decomposition(:,:,1)=[LLr,LHr;HLr,HHr];
First_Level_Decomposition(:,:,2)=[LLg,LHg;HLg,HHg];
First_Level_Decomposition(:,:,3)=[LLb,LHb;HLb,HHb];
First_Level_Decomposition=uint8(First_Level_Decomposition);

%Display Image
subplot(1,2,1);imshow(Input_Image);title('Input Image');
subplot(1,2,2);imshow(First_Level_Decomposition,[]);title('First Level Decomposition');

[Vidya P]

i need the exact inverse DWT code for the above code. can you please provide that??

[Walter Roberson]

There have been quite a number of posts showing dwt2 and reconstruction.

Answer 1

[Ajith Kumar]

if true
%Read Input Image
Input_Image=imread('D:/mytest.png');
%Red Component of Colour Image
Red_Input_Image=Input_Image(:,:,1);
%Green Component of Colour Image
Green_Input_Image=Input_Image(:,:,2);
%Blue Component of Colour Image
Blue_Input_Image=Input_Image(:,:,3);
%Apply Two Dimensional Discrete Wavelet Transform
[LLr,LHr,HLr,HHr]=dwt2(Red_Input_Image,'haar');
[LLg,LHg,HLg,HHg]=dwt2(Green_Input_Image,'haar');
[LLb,LHb,HLb,HHb]=dwt2(Blue_Input_Image,'haar');
First_Level_Decomposition(:,:,1)=idwt2(LLr,LHr,HLr,HHr,'haar',size(Input_Image));
First_Level_Decomposition(:,:,2)=idwt2(LLg,LHg,HLg,HHg,'haar',size(Input_Image));
First_Level_Decomposition(:,:,3)=idwt2(LLb,LHb,HLb,HHb,'haar',size(Input_Image));
First_Level_Decomposition=uint8(First_Level_Decomposition);
%Display Image
figure;
subplot(1,2,1);imshow(Input_Image);title('Input Image');
subplot(1,2,2);imshow(First_Level_Decomposition,[]);title('Reconstructed image');
end

Answer 2

[Ankita Bansal]

Inverse DWT code

Hi, I am assuming that you are using same syntax for coefficients as used in answer given above for calculating DWT. You can use idwt2 function available in MATLAB as following:

if true
% Reconstruction of signal
  Red_Input_Image_reconstructed = uint8(idwt2(LLr,LHr,HLr,HHr,'haar'));  
  Green_Input_Image_reconstructed = uint8(idwt2(LLg,LHg,HLg,HHg,'haar')); 
  Blue_Input_Image_reconstructed = uint8(idwt2(LLb,LHb,HLb,HHb,'haar'));
% Checking whether data obtained after reconstruction is correct or not 
  x=[Red_Input_Image_reconstructed-Red_Input_Image]; 
  y=[Green_Input_Image_reconstructed-Green_Input_Image]; 
  z=[Blue_Input_Image_reconstructed-Blue_Input_Image];  
  max(max(x)) 
  max(max(y)) 
  max(max(z))
% Output Image after reconstruction 
  Output_Image(:,:,1)=Red_Input_Image_reconstructed; 
  Output_Image(:,:,2)=Green_Input_Image_reconstructed; 
  Output_Image(:,:,3)=Blue_Input_Image_reconstructed; 
% Display Image after reconstruction 
  figure; imshow(Output_Image,[]);title('Reconstructed Image');
end

Instead of using “haar” or any other set of filter coefficients available in MATLAB, you can use your own set of filter coefficients for low pass and high pass filters for calculating Wavelet transform and Inverse Wavelet transform. But while calculating Inverse Wavelet transform you must use coefficients corresponding to the coefficients used at the time of decomposition.