Image Quality Enhancement of Digital Mammograms Through Hybrid Filter and Contrast Enhancement

Digital mammogram images contain many noises (i.e., salt and pepper, speckles) and artifacts (i.e., opacity, markers, chest wall, date, and background). Artifacts, illumination, and fewer pixel resolutions lead to poor-quality images, reducing segmentation and classification accuracy. This work aims to develop an improved preprocessing method for breast cancer detection that can be achieved by integrating appropriate noise reduction and contrast enhancement methods. A hybrid filter (HF), consisting of an improved wavelet filter and curvelet filter, is used for noise and artifact removal, respectively. A pixel-based bilinear interpolation (PBI) algorithm is used for image scaling that changes the pixel information of the image. The resized images are passed through a contrast enhancement process to increase the contrast of the images and get a better view. The election-based optimization (EO) algorithm is used to improve the contrast value, which optimally adjusts the gamma intensity and enhances the quality of the images. Contrast enhancement doubles mammogram image quality after noise reduction. The proposed method is evaluated with two datasets, i.e. collected and digital database for screening mammography (DDSM) using the peak signal-to-noise ratio (PSNR) and mean square error (MSE) as parameters. It is found that the MSE value for the mammogram images is reduced using the proposed image enhancement technique. This reduction in MSE leads to an increase in PSNR and improves the image quality of mammograms.