Segmentation of bright and faint circles of an image

Is there any way in MATLAB to segment bright and faint circles of the image shared below?

https://www.dropbox.com/sh/rthfy269nigvydz/at0Qr79bFi

Answer 1

[Image Analyst]

Are they in known, fixed locations? Or do they move around? If they move around, why do they move? Can't you fix the locations and camera?

[Ronaldo]

I do not exactly understand why it is important to know whether the circles are moving or not. To make the question more clear, I have an image from which I want to keep the bright and faint circles and remove other parts of the image. Anyway thanks for your comment.

[Ronaldo]

The bright circles and faint circles are fixed!.

[Image Analyst]

That is very good! You probably have a jig to hold your sample in a reproducible location, and the camera mounted permanently on a solid bracket. Excellent! It means you can use a fixed mask. The reason that is good is because the faint circles are very noisy, low contrast, and will be hard to find automatically. If you know exactly where to look, then the battle is nearly over. It is so much easier than if you had to automatically find the circles.

Now all you have to do is to create a binary image with white where the spots are and black everywhere else. Then call regionprops() passing in both the binary image mask and the gray scale image, and ask for MeanIntensity or whatever else you want to measure.

[Ronaldo]

It seems that my previous comments mislead you a bit. I have to automatically find both bright and faint images (I am not aware of their locations in the image).

[Image Analyst]

Can you increase the exposure time or do something to make the images less noisy?

[Ronaldo]

Unfortunately, No. This is the optimum condition to take the images.