and the background subtracted. The image is normalised based on a maximum and minimum value. This gives a better contrast between the gray backdrop and the droplets.

The hough transform gives the location and diameter of circles found in the image Note that not all the drops are detected