Applying Individual Thresholds to Objects Within the Same Image

I'm analyzing images of cellular nuclei and I am trying to figure a way to remove pixels under a certain intensity within this image. But instead of setting one threshold for the entire image that applies to all the nuclei,I would like to apply my own threshold for each individual nucleus within the same image. Does anybody have an idea to tackle this problem? I attached an example of an image that I am working with.

Answer 1

[Image Analyst]

There is no image attached. What do you mean by "remove", knowing that an image must remain rectangular and you can't have any "missing" pixels which would cause some rows or columns to have a different number of pixels. Do you mean crop out to extract those dark regions to small, separate mini-images?

[Image Analyst]

See my image segmentation tutorial, where I crop out things from the image. http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

Answer 2

[Francisco]

Answer 3

[Francisco]

I apologize about the picture it is now attached. When I set a range of pixels that I want to display based on my cutoff I use

imshow(image, [low value, high value]).

I'm assuming with this command only pixels above the low threshold - infinity are displayed. I have also attached a picture of the frequency distribution of all pixel intensity of one nucleus. The line represents the cutoff that is calculated for the object, and pixels to the right of the line is what I want to keep (note: for the graph the x axis is the pixel intensity value and the y axis is the count at the x value for that point). I know there are better ways to do this but what I currently do is crop each individual nucleus from the entire image and calculate the threshold. I would then like to apply the threshold determined for each individual nuclei to the picture which contains all of the nuclei.