An Algorithm for Counting Microorganisms in Digital Images

Counting microorganisms is one of the main and most common actions in biological activities, including research and clinical analyses. In most cases, such counting procedure is performed manually, in a process that is often lengthy and tedious. As a response to that situation, many methods to automate such a process have been proposed in the literature, but virtually all of them have very strict limitations in their applicability, being effective only in very specific situations. This paper presents an algorithm for cell counting that aims to be more generic than its predecessors. To achieve such a goal, it requires a quick and easy input from the user, which is used to guide the identification and counting process. The resulting algorithm is robust to most situations found in practice, with a drop in performance occurring only when the objects are either strongly clustered or have large regions with the same shade as the background.