Detection of malaria parasites in thin red blood smear using a segmentation approach with U-Net

Malaria is infected by Plasmodium parasite which was a plague in many countries in the world. An alternative way to decrease the number of deaths caused by malaria parasites was that conducting advanced diagnostic for detecting malaria parasite. However, this diagnostic required expert skills and was inclined to human errors. In this situation, automation of this process could greatly help lot of laboratory to handle the disease. In this paper, we presented an automated way to segment the Plasmodium parasites infected in red blood cells using U-net. We conducted our training approach in the three different color spaces which are RGB, HSV and GGB. The proposed method achieved accuracy of 0.9940, 0.9936 and 0.9947 in RGB, HSV and GGB color space respectively. © 2020 IEEE.