Image analysis: identify circle, calculate pixel gradient across image within circle

Hello experts!

I have images I'm collecting from field data. In Matlab, I'd like to:

1. Identify the circle within the image as best as possible
2. Figure out a way to find the differences in pixel values throughout the image within the circle.

Images are large (.tif files, > 2500 x 2500) because I need the resolution, so my attempt at the usual imfindcircles() is difficult and slow. I've attached JPEG version here. Might be able to get similar results with JPEG...

Anyway, the goal is to be able to identify how uniform the color is across the circle. So, in the test image, you can see there are darker areas and lighter areas within the circle. Need help on to quantify that. An ideal image would show the color in the circle to be uniform. In the original images (there will be many), the colors with be different. Some might be more red or purple or brown...whatever. The background will ALWAYS be bright white (it's snow).

I'm thinking that when I take an image, I should have one of those photographer's grayscales to balance the exposure, so each image has the same exposure...? I don't know. I'm not sure where to go from here. Any help will be greatly appreciated.

Thanks!

EDIT:

I found user Image Analyst's demo on Image Segmentation and have begun to adapt that to my needs. I can now

1. Original image, BW histogram, and prompt user to select proper threshold value
2. identify the centroid of the different color "blob"
3. identify the outline of the pixels that need to be analyzed using bwboundaries()

So, now the question how to gather those pixels and the proper way to analyze them. Any ideas?

One idea I have is to gather the mean and standard dev. of the pixels within the identified area. Use these statistics to create a normal distr. plot and measure the full width, half max. of that plot. An image with a more uniform color within the blob will have a different statistics and FWHM, right? Any help would be appreciated.