Analysing the structure of medical images with morphological size distributions

Abstract

The morphological size distribution is introduced as a general concept for analyzing structures in binary as well as in grey-value images. The morphological transformation can be used to treat light and dark structures separately. Thus, it is possible to describe the objects in a configuration and the spaces between or convex and concave parts in analyzing the shape. Examples are given which demonstrate that, using the size distribution, it is possible to characterize texture, shape, and configuration.<>