Segment color dicom image

Hi,

How can i segment colored images like this one, is the rgb threshold is a solution

Answer 1

[John BG]

Bonjour Sbei

1.- Separate the components

A=imread('image_segmentation.jpg')
Ar=A(:,:,1);Ag=A(:,:,2);Ab=A(:,:,3)

2.- Apply the BW functions you show in the question to each primary colour

3.- Find the intersection pixels of the results found in the 3 layers

If you find this answer of any help solving this question, please click on the thumbs-up vote link,

thanks in advance

John

[e-mail]

Answer 2

[Image Analyst]

It looks like someone has already segmented it. The image started out as gray scale and then somehow segmented it and produced a pseudocolored version of it.

I'm not sure what you want to find here. You can see from the color gamut below that there are not well clustered colors. Everything is kind of blended.

As you rotate that gamut around, you get some very strange looking shapes, like "sheets" in the RGB Gamut space.

And here is what it looks like if you look straight down the V axis in HSV color space:

What color(s) do you want to find?

Check out my Color Segmentation tutorials in my File Exchange: http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

[sbei arafet]

Hi, Thanks for your answer What color(s) do you want to find? as i'm working with fuzzy connectedness , my object is to segment tumor regions , since hybrid images are not in gray space like PET-MRI , so the program will be semi-automatic , i will give seeds in tumor region and the algorithm will determine all the tumor region, so i'm not working for a specific color, thanks for the link i will check it.

Regards

[Zubda Butt]

Can you please explain your input arguments I and A in affinity function.

[Shaveta Arora]

can I get the image database

[Walter Roberson]

Which image database, Shaveta Arora ?