A Study of Dengue Infection Segmentation, Feature Extraction and Classification

Abstract

Aedesaegypti mosquito spared the dengue viral illnesses. The world�s greatest developing outbreak is dengue fever. Day �by day the rate of dengue has become significantly around the globe increases. Dengue infections are of three forms: Dengue fever additionally perceived as �break bone� fever, Dengue Haemorrhagic Fever (DHF), Dengue Shock Syndrome (DSS) which are life debilitating. Doctors need to capture approximately 20 to 50 pictures of white blood cell from different angle to identify the disease. The platelet count is estimated using various segmentation techniques and morphological operations with the help of the platelets count dengue fever infection is �detected. A technique used for segmentation are mainly thresholding based that is not segment exact part of defected platelet. But, the result was not so efficient in providing the spatial detail information of the actual disease part. So here we are going to use Fuzzy based algorithm to segment WBC Platelets. There are different feature extraction methods are apply platelet are size, shape and area. But it was not giving the exact results. So here we are going to use Haarlick Features for WBC platelets. And any machine learning method SVM, ANN, Decision Tree will be used for the classification of dengue infection types.