Cervical Cell Segmentation from Overlapped Cells using Fuzzy C-Means Clustering

Abstract

Cervical cancer is the symptomless disease to cause death amongst women due to cancer. Most of the cervical cancer diagnosis process microscopic images are taken as sample to identify Segmentation of cervical cells. In this paper, Fuzzy c-means clustering algorithm is used to preserve the colour and data loss during segmentation is minimal. It accurately segments the individual cytoplasm and nuclei from a cluster of overlapping cervical cells. Recent methods cannot undertake such absolute segmentation due to various challenges involved in delineating cells coping with overlap and poor contrast. Improved method for detecting overlapping cervical cells using advanced tests yields better results in detection. The cervical cancer can be prevented through both early detection and best treatment based on the acuteness of the disease.