Features Extraction of Mamographic Image using Zoning Method

Mammographic images are images that can be used as a tool to detect the presence of breast cancer. The presence of the disease is shown in the characteristics of the breast tumor object that appears in mammographic images. Visually, the structure of mammographic images has abnormalities so it is difficult to recognize the characteristics of these images. Therefore, in this paper an algorithm is used to extract the features of the breast tumor shape that appears in mammographic images using of zoning method. The zoning method divides the image of mammographic characters into several parts, then special features in each part are extracted for producing an eigen value that corresponds to the largest eigen value. Then, the eigen values are used to form the main characteristic space of the mammographic image. The results show that the algorithm is able to localize the suspected area as a form of breast tumor and can produce the characteristics of the images. Furthermore, the features produced in this paper can be used in the process of classifying breast tumor disease.