Hybrid Gabor based Local Binary Patterns Texture Features for classification of Breast Mammograms

Breast cancer is one of the most widespread cancers in the world. A digital mammogram is an X-ray picture of the breast, which used for early detection of breast cancer and other abnormalities. The early detection of the cancer can reduce mortality. Mammograms help radiologists to detect the malignant masses precisely at their early stage. But, the mammograms are usually characterized by low contrast therefore it is not easy to read by the radiologist to detect small masses and micro-calcifications, which are indirect signs of malignancy, precisely and in their early stage. So we need a Computer aided diagnosis (CAD) system for mammography, which simulates the process of the radiologist and used to interpret mammography image and check for the presence of breast cancer and distinguish between malignant and benign tumors of breast cancer. In this paper, we introduce an approach for mammogram breast cancer detection system. This system involves three phases. : preprocessing of mammogram image, features extraction, and classification. Texture features based upon Gabor filters with Local Binary Patterns has been extracted for classification.