Computer aided diagnosis of Malaria disease for thin and thick blood smear microscopic images

Malaria Is a serious health Issue and causes a million deaths in a year globally. The present gold standard of malaria diagnosis, recommended by world health organization (WHO) is the manual microscopy method of Giemsa-stained blood smears, which is a laborious process requiring expert technicians. This paper presents a robust and fast algorithm that identifies Malaria parasites from both thin and thick blood smears. In the proposed method, first the images are pre-processed in order to remove noise variations occurs due to microscope lenses and different lighting conditions. This step makes the proposed algorithm more robust to those noise conditions. Next to this, image segmentation is performed using histogram based adaptive thresholding followed by mathematical morphological operations. The segmented images are used further to detect infected red blood cell (RBC) by the malaria parasite. The detection of infected RBC is done using unsupervised learning technique, which makes the detection process faster. The proposed algorithm is experimented on the different images and it shows that the proposed algorithm is suitable for computer aided diagnosis (CAD) of Malaria disease.