Measure the surface area of holes in cell monolayer


I have some questions concerning the processing of my microscope images. I have a lot of images of RPE cell monolayers. In the two attached images you can see, that there are holes in the cell layer, which look like little bubbles.
I would like to determine the surface area of the holes. Since I have only limited experience with imageJ, I did not find a proper solution. I tried already the common methods, which are also used to count cells (thresholding, particle analyser, edge detection, weka segmentation, etc.)
On major problem is the variation in pigmentation of the cells. As you can see in the example images, the intensity of the cell‘s pigmentation differs quite a lot. Some of the unpigmented areas are also detected as holes. It seems to be difficult to find one workflow for all the images.
It would be great, if anyone had another idea or recommendation for me.
Thank you very much in advance!