Journal of Contemporary Issues in Business and Government

Abstract

Cervical cancer is the foremost cause of death among women across the globe due to cancer. Exact and error-free detection can save lives. Earlier detection of Cervical Cancer was done by a microscopic smear test based on the calculation of parameters of the cell nucleus of the sample such as its shape and size as smear is analyzed to microscope is an extremely challenging task. Hence, Digital Image Processing technique are expected to identify abnormalities in human cell. Consequently, a comprehensive machine learning technique has been proposed in this paper. The proposed technique gives the features and shape of cytoplasm, nucleus in the cervix cell. KNN and SVM are trained with the features and shape of the segmented cell and compared with unknown cervix cell sample with this technique. The accuracy rate of 86% for SVM and 70% for GLCM is achieved.