A Novel Noise Removal in Digital Mammograms based on Statistical Algorithms

Abstract

The noise removal is being a substantial phase for the computer-assisted detection (CAD) based breast cancer diagnosis using mammogram medical images. A proficient method for the salt-and-pepper or impulse noise eradication in digital mammograms is implemented. The approach depends on the statistical measures like mean, median and standard deviation quantities. This calculates the new intensity which is to be substituted in the impulse area by determining those measures in neighbour points of the taken mammogram images. The proposed is simply an iterative method that aims to take away the salt and pepper otherwise impulse noise devoid of affecting the boundaries and other major significant portions of the image. The approach is compared with several existing methods and it provides enhanced noise removal performance over others.