A Study of Mammographic Image Segmentation with its Morphological Operation

Abstract

. The Malign cell extraction and segmentation differentiation from normal cells is widely researched topic. The process of segmentation with single strategy might miss the features leading to increased mortality rate. This work characterizes the different segmentation methods and two simulation tools for mammogram images. The non-feature pixel values are represented by the nearest feature pixel in distance by watershed segmentation. Simulations are performed with ImageJ using Morphological library where binary mammogram images are analysed with connected components and distance based watershed transform. Finally the mammogram image in DICOM format is analysed for segmenting spanning lower and upper threshold with clustering..