HAMMF: A Hybrid Adaptive Modified Median Filter for Denoising Mammogram

Abstract

Mammography is crucial for detecting breast cancer, but noise can hinder accurate interpretation. This paper proposes a two-stage hybrid de-noising algorithm for digital mammograms. First, the Adaptive Modified Median Filter (AMMF) removes noise, and then Guided Filter (GF) is applied to preserve edges and details. Experiments on Mammographic Image Analysis Society (MIAS) dataset show the filter effectively removes salt-and pepper noise. Performance is evaluated using metrics like Mean Square Error (MSE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index Measure (SSIM), comparing it to other filters like Median, Adaptive Median, Hybrid Median, Bilateral, and GF. The proposed method effectively suppresses noise, preserves edges, and improves visual quality.