Haralick fetaures based mammogram classification system

The second cause of the death among women arises due to breast cancer that affects the breast tissues. The efficient prognosis way of breast cancer is processed with the aid of mammogram images. The proposed mammogram classification system improves the diagnosis and early detection of breast cancer by using mammogram images. It helps radiologists to diagnose cancer accurately. MIAS database images are used for the evaluation. Thirteen Haralick texture features such as correlation, contrast, entropy, homogeneity and energy are extracted. The robust k-nearest neighbor (KNN) is used as classifier, and it classifies the mammogram images into two categories, which are normal and abnormal. The proposed approach provides satisfactory classification accuracy of over 92%.