Improved mean shift Clustering based MRI image segmentation for Cervical cancer
Abstract
Cervical cancer unfortunately has emerged as one of the major causes of death, consequently making its detection extremely vital, although very challenging to diagnose and treat. The present study deals mainly with the performance study and analysis of modified mean shift Clustering (MMSC) techniques suitable for cervical images segmentation. This work presents the design, implementation and evaluation of a cervical images segmentation system employing MMSC for improving cervical cancer classification using MRI images. The input image is initially preprocessed with Gaussian filter and canny edge detection then the MMSC approach is utilized to segment the cervical cancer region. The proposed methodology is evaluated with necessary performance metrics like RMSE and PSNR.