In computer aided mammography algorithms there are several processing steps, which must be performed. The basic segmentation procedure involves extracting the principal feature on a mammogram; the breast border. This is performed by segmenting the breast and the non-breast into distinct regions. In this paper, a method for extracting the breast border is proposed. The method has performance similar to established techniques but with higher degrees of automatization and robustness. It iteratively adapts a model of the background to ensure a robust object detection yielding a smooth outline of the breast. The main idea is to identify the "knee" in the cumulative intensity histogram of the image. The intensity value
at the knee is thereafter used to automatically define a region, to be modelled by a two-dimensional polynomial surface of degree two. The modelled background is then subtracted from the original image. The procedure described is iteratively performed until the degree of non-uniformity of the grey-scale background is smaller then a certain value. Thereafter the difference image is post-processed by a flood-filling algorithm, a new threshold is estimated as above and applied to yield a binary image. Lastly morphological operations are performed to smoothen the breast border. In conclusion, the strength in the proposed method, compared to similar methods, is that it makes use of an iterative approach to reduce the effects of the background, it produces smooth edges and automatically finds
thresholds. It is also evaluated on the entire MIAS database (322 images) with a performance of 94%.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.