correcting x-ray image

hello... i have the following code to correct the contrast of x-ray image,, but at last it gives a black image not x-ray,, Can any body help ???

load('heel_effect.mat')
imread('Hand.png')
subplot 231 
imshow('Hand.png') 
title('Origin')
tissue=tissue_eq_slab_image ; 
h=x_ray_HeelEffect_image;
subplot 232
imagesc(tissue); colormap gray
title('Tissue Slab')
subplot 233
imagesc(h); colormap gray
title('Heel effect')
[M,N]=size(tissue);
mid_line=tissue(M/2,:);
subplot 234
imshow(mid_line)
title('Mid Line')
[m,n]=size(mid_line);
for c=1:n
    T(c)=mid_line(:,c);
end
subplot 235
plot(1:n,T)
title('Relation')
log_T=log (double(T+eps));
log_c=log(double((1:n)+eps));
% I=A*x^Gama
%Ln I=Ln A+ Gama Ln x
 z=polyfit (log_c, log_T,1);
 Gama=z(1,1); %slope= Gama
 A=exp(z(1,2)); %constant=Ln A
 [m1,n1]=size(h);

 for i=1:n1
    corrected(:,n1)=tissue(:,n1)/A*(n1.^Gama);
 end

subplot 236 
imagesc(corrected)
title('Corrected Image')

[sixwwwwww]

Can you please attach the image? Can you please format your code to be displayed correctly using "{} Code"?

[Christine Ak]

load('heel_effect.mat')
imread('Hand.png')
subplot 231 
imshow('Hand.png') 
title('Origin')
tissue=tissue_eq_slab_image ; 
h=x_ray_HeelEffect_image;
subplot 232
imagesc(tissue); colormap gray
title('Tissue Slab')
subplot 233
imagesc(h); colormap gray
title('Heel effect')
[M,N]=size(tissue);
mid_line=tissue(M/2,:);
subplot 234
imshow(mid_line)
title('Mid Line')
[m,n]=size(mid_line);
for c=1:n
    T(c)=mid_line(:,c);
end
subplot 235
plot(1:n,T)
title('Relation')
log_T=log (double(T+eps));
log_c=log(double((1:n)+eps));
% I=A*x^Gama
%Ln I=Ln A+ Gama Ln x
 z=polyfit (log_c, log_T,1);
 Gama=z(1,1); %slope= Gama
 A=exp(z(1,2)); %constant=Ln A
 [m1,n1]=size(h);

   for i=1:n1
      corrected(:,n1)=tissue(:,n1)/A*(n1.^Gama);
   end

    subplot 236 
  imagesc(corrected)
  title('Corrected Image')

[sixwwwwww]

Dear Christine, "heel_effect.mat" is not attached yet and also can you please explain what is meaning of these lines of your code?

tissue=tissue_eq_slab_image; 
h=x_ray_HeelEffect_image;

because variables "tissue_eq_slab_image" and "x_ray_HeelEffect_image" are not defined anywhere. So I assume that you mean as follows

tissue_eq_slab_image = imread('Hand.png')
x_ray_HeelEffect_image = load('heel_effect.mat')

Is it correct?

[Christine Ak]

Thank U 4 replying

Those variables are in heel_effect.mat file which i can't attach it here !

the problem is not here

but i don't know what is the error exactly :(

[sixwwwwww]

what is size of your ".mat" file?

[Image Analyst]

We can't really run your code unless you attach the mat file - you have to zip it up for now (until Answers gets that issue fixed). Have you actually observed values of corrected in the variable editor and verified that all values are exactly zero? Or does it just look black when you display it?

[Christine Ak]

160 K I've submitted the File heel_effect.mat

[Image Analyst]

Nothing is attached.

[Christine Ak]

i can't attach it here !!

[Image Analyst]

zip up the mat file and click the paper clip icon when you make a comment!!

[Wesam Mohamed]

Is this the full answer with run ?!

Answer 1

[Youssef Khmou]

You can use the histogram equalization , see this example :

 I = imread('tire.tif');
 J = histeq(I);
 imshow(I)
 figure, imshow(J)

see

 doc histeq

[Christine Ak]

Thxx 4 Replying but the image i have is 3 Dimensional...."" histeq"" function gives error :(

[Youssef Khmou]

Apply hiseq for each channel : example :

 I=im2double(imread('autumn.tif'));
 X=zeros(size(I));
 for n=1:3
      X(:,:,n)=histeq(I(:,:,n));
 end

[Image Analyst]

No offense, but do you like what histeq does? I find it gives unrealistic and unnatural looking images. Even a primitive contrast stretch like imadjust() will give more pleasing images than histeq(). I think histeq is just in there for instructional purposes but I doubt anyone in real world situations considers it a good thing to do. A gamma like she tried will probably look okay but that just adjusts intensity. Some of the better filters do that plus do spatial frequency enhancements to see details better.

[Youssef Khmou]

I agree, but it may work for X ray images, waiting for her answer..

[Christine Ak]

actually it doesn't work with X-rays...... My purpose is to enhance the contrast in the heel effect of the image the problem in the image is that the light distributed ((not)) equally at the sides

[Christine Ak]

Thankk You Guys 4 ur help... I found out the error it was in the second for loop...it must be

   for i=1:n1
      corrected(:,i)=h(:,i)/(A*(i.^Gama));
   end