White blood cell recognition via geometric features and naïve  bays classifier

Blood assessments are of the maximum crucial and frequently asked medical examinations. A manual microscopic evaluation must be done while a blood pattern is suspicious of abnormality. This manual technique is tedious, time ingesting and subjective. Automating microscopic blood type is appropriate to assist the pathologists to hurry-up and induce the consequences accuracy. Segmentation is the primary and common step in computerized WBCs category. On this paper, have been presented a powerful method for automated WBCs nuclei segmentation. The technique is based on gray scale contrast enhancement and then using Otsu thresholding tech-nique to segment WBCs. There are four features have found to extract the data from the segmented image. These features are (Area, Perim-eter, diameter, Circularity. Then these data was classified using Naive Bayes classifier under weka program. The approach is examined on 260 blood pictures. The class overall performance is quantitatively evaluated at the take a look at set to be 97,1 %. This overall performance is excessive in comparison to other related work done at the identical dataset.