ABSTRACT:
Malaria is a life-threatening disease caused by parasites that are transmitted to people through the bite of female anopheles mosquito. Malaria is usually found in tropical and subtropical climates where the parasites that cause it live. Once the parasite enters the human body, it lodges itself in the liver where it multiplies approximately 10,000 times. The diagnosis of malaria is done by various methods which include thick and thin smear method, rapid diagnostic test, antigen detection etc. The limitations of these methods includes false positive rate. This paper presents a methodology to identify the malarial parasites so that the false positive rate is reduced. The color based discrimination is one of the most feasible methods as it shows greater accuracy and efficiency. The K-means clustering technique is followed for color based discrimination. The segmented malarial cell from K-means clustering is further subjected to morphological operations to extract the feature for classification. The results show that the proposed method helps in the detection of infected red blood cells thus improving decision-making for malaria diagnosis.