Efficient preprocessing filters and mass segmentation techniques for mammogram images

Abstract

Breast tumor is a champion among the most broadly perceived sort of development among women that creates from breast tissue. Still the correct reason for the breast disease stays obscure. Early discovery and determination is the best methodology to control the tumor movement. Mammography is the right now suggested imaging strategy for early assurance and determination of breast danger. A mammogram can distinguish unusual areas in the breast that resemble a malignancy. Mammogram pictures are observed to be hard to decipher so a CAD is turning into an undeniably essential device to help radiologist in the mammographic lesion interpretation. Preprocessing was considered as an essential stride in mammogram picture investigation. Exactness of preprocessing will decide the achievement of the rest of the procedure, for example, segmentation, classification and so forth. In this paper, mean, median, adaptive median, gaussian and wiener denoising filters are utilized to evacuate salt and pepper, speckle and gaussian noises from a mammogram picture and these filters were looked at in light of the parameters, for example, PSNR, MSE and SNR to figure out which filter is better to remove these noises in mammogram pictures. The segmentation is a process in which changing the representation of an image such that it is easier to analyse. This paper compares various mass segmentation techniques used in mammogram images.