Morphological Image Processing for Evaluating Malaria Disease Cecilia Di Ruberto 1 , Andrew Dempster 2 , Shahid Khan 3 , and Bill Jarra 3 1 Department of Mathematics, University of Cagliari, Cagliari, Italy dirubert@vaxca1.unica.it 2 Department of Electronic Systems, University of Westminster, London, UK dempsta@cmsa.wmin.ac.uk 3 National Institute for Medical Research, London, UK Abstract. This work describes a system for detecting and classifying malaria parasites in images of Giemsa stained blood slides in order to evaluate the parasitaemia of the blood. The first aim of our system is to detect the parasites by means of an automatic thresholding based on a morphological approach. Then we propose a morphological method to cell image segmentation based on grey scale granulometries and openings with disk-shaped elements, flat and hemispherical, that is more accurate than the classical watershed-based algorithm. The last step of the system is classifying the parasites by morphological skeleton. 1 Introduction In malarial blood the red corpuscles of vertebrates are infected by malaria para- sites. The parasite develops in a highly regulated manner through distinct cycles in the vertebrate host [8]. The parasite attacks red corpuscles, in which it first appears as minute speck of chromatin surrounded by scanty protoplasm, and gradually becomes ring-shaped and is known as a ring or immature trophozoite. It grows at the expense of the red cell and assumes a form differing widely with the species but usually exhibiting active pseudopodia, i.e. projections of the nuclei. Pigment granules appear early in the growth phase and the parasite is known as a mature trophozoite. As the nucleus begins to divide and take up peripheral positions, the parasite is known as a schizont. The infected red blood cell ruptures. Some parasites on entering red cells become round sexual gameto- cytes, instead of asexual schizonts. The aim of our system is to detect the parasites using a scan of a colour photo- graph of stained malarial rodent blood from a microscope in order to evaluate the parasitaemia of the blood i.e. counting the number of parasites per number of red blood cells [1]. A manual analysis of slides is tiring, time-consuming and requires a trained operator. So our task is to automate the process. The image processing system is made of three main steps: detection of parasites, cell seg- mentation and classification of parasites. In Section 2 we describe the different phases of the image analysis, beginning from parasites detection. We propose a method to automatically separate the parasites from the rest of an infected C. Arcelli et al. (Eds.): IWVF4, LNCS 2059, pp. 739–748, 2001. c  Springer-Verlag Berlin Heidelberg 2001