Detecting and Counting of Blood Cells Using Watershed Transform: An Improved Methodology

Abstract

The necessity for practicality, reliability, and agility has stimulated the creation of new tools for the health area. The blood cell counting is an important exam that helps to diagnose various diseases besides to be an important field of study in biomedical engineering. Currently, the red blood cells count (RBC) and white blood cells (WBC) have been performed by the automated method. However, still does not dispense the use of the manual method, through the preparation of blood smears, in specific cases or only for confirmation of diagnostics. In the last years, MATLAB® software has been used to develop algorithms to facilitate cell counting. Thus, the present work implements an algorithm based on Watershed transform, where its results show a better performance of 34% in execution time in comparison with other works developed, as well as a computational performance of 1.98 seconds, as also showing high reliability in your digital count.