What is the best method for analyzing bacteria?

I'm trying to track the position of bacteria over generations and have attempted two methods of doing so. The first involves large amounts of adjusting image processing parameters such as strel which has been successful but takes along time. I have also been attempting to use statistical methods such as a using histograms to get rid of the background however I am having some issues with locating the correct threshold to use in im2bw. If anyone could give me some pointers I would be very grateful! One of the images I am using is attached

Answer 1

[Image Analyst]

I don't know why determining parameters takes such a long time. Anyway, why can't you just threshold and find the centroid?

Answer 2

[Nathan Costin]

That is what I have done in the past however it requires other morphological processes that the designer has to change. I am hoping to make an automatic code so having a method that gets rid of the most amount of noise to a high precision would be useful.