Fuzzy c-means Segmentation on Enhanced Mammograms Using CLAHE and Fourth Order Complex Diffusion

Abstract

Cancer statistics all around the globe are rising day by day, out of which breast cancer is the dominating one in women. Mammography is used to detect the presence of cancerous cells and computer-aided detection technologies are used to get more accurate results. There are different image processing techniques which are applied for cancer detection in mammograms. In this work, a combined methodology is used for enhancing the contrast of the mammogram which includes contrast limited histogram equalization (CLAHE), morphological gradient and fourth order nonlinear complex diffusion based unsharpening. The enhanced mammogram is then segmented using fuzzy c-means. An analysis is done on each of the following cases with Fuzzy-c means segmentation: CLAHE enhanced, morphologicall gradient enhanced, fourth order nonlinear complex diffusion enhanced and combined enhanced image using the above methods.