Unsupervised morphological granulometric texture segmentation of digital mammograms

Segmentation via morphological granulometric features is based on fitting s tructuring elements into image topography from below and above. Each struct uring element captures a specific texture content This paper applies granul ometric segmentation to digitized mammograms in an unsupervised framework, Granulometries based on a number of flat and nonflat structuring elements a re computed local sire distributions are tabulated at each pixel, granulome tric-moment features are derived from these size distributions to produce a feature vector at each pixel, the Karhunen-Loeve transform is applied for feature reduction, and Voronoi-based clustering is performed on the reduced Karhunen-Loeve feature set Various algorithmic choices are considered, inc luding window size and shape, number of clusters, and type of structuring e lements. The algorithm is applied using only granulometric texture features , using gray-scale intensity along with the texture features, and on a comp ressed mammogram. Segmentation results are clinically evaluated to determin e the algorithm structure that best accords to an expert radiologist's view of a set of mammograms. (C) 1999 SPIE and IS&T. [S1017-9909(99)00901-0].