Mammogram Image Grade Gauging of Denoising Filters & Enhancement Methods

Abstract

Breast cancer is the most treacherous tumour among women and its early detection enhances the chances of survival of the patient. Screening mammography improves a physician's ability to detect even small tumours which cannot be felt physically by the patient. Mammographic image noises influence the diagnostic images which can affect the diagnostic process. Hence it is indispensable to filter out the noises by preserving important features of the image. This paper investigates and identifies the most appropriate denoising filter and enhancement technique among mean, median, adaptive median, gaussian, wiener, contrast stretching, histogram equalization and Contrast Limited Adaptive Histogram Equalization (CLAHE). The matrices used to analyse the performance is Mean Square Error (MSE), Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index Metric (SSIM). From experimental results and analysis, it is proved that adaptive median filter and histogram equalization techniques are efficacious in removing noise and thereby enhancing the calibre of the image.