How can i do BI-HISTOGRAM EQUALIZATION in MATLAB

The folling program i tried but it didn't work properly.

I=imread('tire.tif');

%performing bi-histogram equalization

L=256;

[m,n]=size(I);

len=m*n;

x=reshape(I,len,1);

xm=round(mean(x));

xl=x([x<=xm]);

xu=x([x>xm]);

%%%%%%%%%%%%%%%%%%%%%%%%%%

xlpdf=hist(xl,[0:xm]);

xlnpdf=xlpdf/length(xl);

xupdf=hist(xu,[xm+1:255]);

xunpdf=xupdf/length(xu);

%%%%%%%%%%%%%%%%%%%%%%%%%

skl=xm*xlnpdf*triu(ones(55));

sku=(xm+1)+((L-1-(xm+1))*xunpdf*triu(ones(201)));

sk=[skl,sku];

y0=zeros(m,n);

for k=0:L-1

        list=find(I==k); %find locates the non-zero values

        y0(list)=sk(k+1);

end

y0=uint8(y0);

imwrite(y0,'biHE.tif','TIFF');

j=imread('biHE.tif');

Answer 1

[Image Analyst]

Use intlut() instead of that for loop. You create a look up table of output values for each input value.

[Tuhin Sheikh]

Sir, i didn't get your answer, could you please help me in detail...

Have i correctly written the algorithm?

Using this program the value of PSNR is less than that of Globally

equalized image, but actually the PSNR of Bi-histogram is to be greater

than histogram equalization!!!

[Image Analyst]

Well I'm not a big fan of histogram equalization. And I really don't see why separately doing histogram equalization on the low and high gray levels would be any good, but I did what you want, or at least what your algorithm tries to do. Here is the code:

clc;    % Clear the command window.
close all;  % Close all figures (except those of imtool.)
imtool close all;  % Close all imtool figures if you have the Image Processing Toolbox.
clear;  % Erase all existing variables. Or clearvars if you want.
workspace;  % Make sure the workspace panel is showing.
format long g;
format compact;
fontSize = 20;

% Check that user has the Image Processing Toolbox installed.
hasIPT = license('test', 'image_toolbox');
if ~hasIPT
  % User does not have the toolbox installed.
  message = sprintf('Sorry, but you do not seem to have the Image Processing Toolbox.\nDo you want to try to continue anyway?');
  reply = questdlg(message, 'Toolbox missing', 'Yes', 'No', 'Yes');
  if strcmpi(reply, 'No')
    % User said No, so exit.
    return;
  end
end

% Read in a standard MATLAB gray scale demo image.
folder = fullfile(matlabroot, '\toolbox\images\imdemos');
button = menu('Use which demo image?', 'CameraMan', 'Moon', 'Eight', 'Coins', 'Pout');
if button == 1
  baseFileName = 'cameraman.tif';
elseif button == 2
  baseFileName = 'moon.tif';
elseif button == 3
  baseFileName = 'eight.tif';
elseif button == 4
  baseFileName = 'coins.png';
else
  baseFileName = 'pout.tif';
end

% Read in a standard MATLAB gray scale demo image.
folder = fullfile(matlabroot, '\toolbox\images\imdemos');
% Get the full filename, with path prepended.
fullFileName = fullfile(folder, baseFileName);
% Check if file exists.
if ~exist(fullFileName, 'file')
  % File doesn't exist -- didn't find it there.  Check the search path for it.
  fullFileName = baseFileName; % No path this time.
  if ~exist(fullFileName, 'file')
    % Still didn't find it.  Alert user.
    errorMessage = sprintf('Error: %s does not exist in the search path folders.', fullFileName);
    uiwait(warndlg(errorMessage));
    return;
  end
end
grayImage = imread(fullFileName);
% Get the dimensions of the image.  
% numberOfColorBands should be = 1.
[rows, columns, numberOfColorBands] = size(grayImage);
if numberOfColorBands > 1
  % It's not really gray scale like we expected - it's color.
  % Convert it to gray scale by taking only the green channel.
  grayImage = grayImage(:, :, 2); % Take green channel.
end
% Display the original gray scale image.
subplot(2, 2, 1);
imshow(grayImage, []);
title('Original Grayscale Image', 'FontSize', fontSize);
% Enlarge figure to full screen.
set(gcf, 'units','normalized','outerposition',[0 0 1 1]);
% Give a name to the title bar.
set(gcf,'name','Demo by ImageAnalyst','numbertitle','off') 

% Calculate mean gray level.
meanGL = mean2(grayImage)

% Let's compute and display the histogram.
[pixelCount, grayLevels] = imhist(grayImage);
subplot(2, 2, 2); 
bar(pixelCount);
grid on;
title('Histogram of original image', 'FontSize', fontSize);
xlim([0 grayLevels(end)]); % Scale x axis manually.
% Put line at mean
yl = ylim;
line([meanGL, meanGL], [yl(1), yl(2)], 'LineWidth', 2, 'Color', 'r');

binaryImageLow = grayImage < meanGL;
binaryImageHigh = grayImage >= meanGL;
% Display the image.
subplot(2, 2, 3);
imshow(binaryImageLow, []);
title('Binary Image Below Mean', 'FontSize', fontSize);
% Display the image.
subplot(2, 2, 4);
imshow(binaryImageHigh, []);
title('Binary Image Above Mean', 'FontSize', fontSize);

% Get maps for each
[~, lowMap] = histeq(grayImage(binaryImageLow));
[~, highMap] = histeq(grayImage(binaryImageHigh));
% Combine them.
intMeanGL = int32(meanGL);
lookupTable = uint8(256*[lowMap(1:intMeanGL), highMap(intMeanGL+1:end)]);

%========================================================
% Use lookup table to map image.
% KEY PART RIGHT HERE!!!
dualHistEqImage = intlut(grayImage, lookupTable);
%========================================================

% Display the image.
figure;
imshow(dualHistEqImage, []);
title('Output Image', 'FontSize', fontSize);
% Enlarge figure to full screen.
set(gcf, 'units','normalized','outerposition',[0 0 1 1]);
% Give a name to the title bar.
set(gcf,'name','Demo by ImageAnalyst','numbertitle','off')

[Tuhin Sheikh]

Wow sir, what i was trying to do for many days and you did it within few

minutes!! Amazing sir and thank you very much.

Sir just explain me these codes

% Get maps for each

[~, lowMap] = histeq(grayImage(binaryImageLow)); what does ~ stand for?

[~, highMap] = histeq(grayImage(binaryImageHigh));

% Combine them.

intMeanGL = int32(meanGL);

lookupTable = uint8(256*[lowMap(1:intMeanGL), highMap(intMeanGL+1:end)]);

[Image Analyst]

~ means to ignore that returned argument. I didn't need that image since it would histogram equalize the whole image and I just wanted it on half the image, which I would do later with intlut().

By the way, histogram equalized images usually look pretty bad and unnatural. I never use them.

[Tuhin Sheikh]

i'm new in this genre sir, i'm just trying to implement the proposed method of

YEONG-TAEG KIM titled "Contrast Enhancement Using Brightness Preserving Bi-Histogram Equalization"

http://st04110083.etu.edu.tr/makale/Contrast%20Enhancement%20Using%20Brightness%20Preserving%20Bi-Histogram%20Equalization.pdf

According to the paper Bi histogram equalization gives out image with better contrast than that of usual global histogram equalization.

But i wonder this program is giving me the output image with with less contrast than global HE based on the measure PSNR( http://en.wikipedia.org/wiki/Peak_signal-to-noise_ratio )

[Image Analyst]

If that's using a global histogram equalization but just using each part fo the histogram independently, then I don't see how it would be any good. But I didn't read the paper so perhaps you didn't describe it correctly. Usually if global histogram equalization is no good, which it often isn't, people will use CLAHE, which is a contrast limited locally adaptive histogram equalization. This can be good for some types of images (e.g. images of pages that you want to do OCR on with huge gradients/shading). This is done by the MATLAB function adapthisteq() in the Image Processing Toolbox.

[afrin sulthana]

can i use this program for contrast enhancement and brightness preserving using bi histogram equalisation

[Image Analyst]

Who are you asking? All I know is the opinions and guesses I gave in my last comment. I don't have any code for bi-histogram equalization.