How can I do the histogram equalization?

Hello

I think that the histogram equalization on the 3 components of a color image has the effect of increasing the contrast in order to facilitate the detection of the region of interest that is darker than the other part of the image. How can I do this using matlab? thanks

Answer 1

[Aneesh Balaji]

Hey For a color image split into 3 components and perform histeq on each of them and then recombine them.

[Tomas]

I used these lines but i got this result

r=im(:,:,1);
v=im(:,:,2);
b=im(:,:,3);
r1 = histeq(r);
v1 = histeq(v);
b1 = histeq(b);
img = cat(3,r1,v1,b1);

http://s12.postimage.org/cqjo80qpp/result.jpg

Answer 2

[Image Analyst]

Using histeq() on individual color channels is not a good method. You will introduce color artifacts which may be severe. You're best off converting to hsv color space, then using histeq() or adapthisteq() on only the v channel, then recombine and transform back to rgb color space.

[Tomas]

I tried this but I got also a bad result

im=rgb2hsv(im);
h=im(:,:,1);
s=im(:,:,2);
v=im(:,:,3);
v1 = histeq(v);
img = cat(3,h,s,v1);
img2=hsv2rgb(img);

[Image Analyst]

Looks like it should be right. Maybe it's some kind of uint8/double conflict. Can you post your image somewhere?

[Tomas]

I'm working on a large number of images. I just need to equalize the histogram to facilitate detection of the dark object. this is an example http://s13.postimage.org/99yr0g2p3/image.jpg

[Image Analyst]

Histogram equalization is not necessary for that purpose. In fact, it's rarely necessary for anything. And it looks like you'd do best with a texture filter, like entropyfilt() or stdfilt(), at least for the one image you posted.