Noise Removal Filtering Methods for Mammogram Breast Images

Abstract

Breast cancer detection in the early stage is an important factor to reduce the mortality rate. Mammogram examination is one of the best optimistic from various approaches used in the early detection of breast cancer at a different stage of cancer and the raw mammogram images are required to pre-process for better radiologist perception and to obtain an enhanced and clear image. It also helps to extract the Region of Interest from the processed image by using statistical feature methods to find the size and shape of the tumor. This paper is on an experimental study performed on sample mammogram images and applies different noise smoothing methods. Methods used to remove noise from the images by applying filtering methods like Gaussian Filter, Tri-State Filter, Mean Filter, Mean-Median Filter, Threshold Filter, Bilateral Filter, Wiener Filter, and Adaptive filter. The processed and obtained quality image will help doctors and radiologists to give an accurate impression on a patient case study. Results: quality of the image obtained on sample mammogram images of CBIS-DDSM dataset achieved min 80% of quality PSNR values.