Segmentation of blood images using morphological operators

This work describes a part of a malarial image processing system for detecting and classifying malaria parasites in images of Giemsa stained blood slides in order to evaluate the parasitaemia of the blood. A major requirement of the system is an efficient method to segment cell images. This paper introduces a morphological approach to cell image segmentation more accurate than the classical watershed-based algorithm. We have applied grey scale granulometries based on opening with disk-shaped elements, flat and non-flat. We have used a non-flat disk-shaped structuring element to enhance the roundness and the compactness of the red cells improving the accuracy of the classical watershed algorithm, while we have used a flat disk-shaped structuring element to separate over lapping cells. These methods make use of knowledge of the red blood cell structure that is not used in existing watershed-based algorithms.