Having discussed the theory of gray-scale mathematical morphology, we now present some procedures that can be employed for various purposes, including filtering, marking, and segmentation.
Various gradients are used in image processing to detect edges, the basic principle being that large gradients represent points where there is a rapid light-to-dark (or dark-to-light) change. Very often these gradients are digital versions of differential gradients. In morphological image processing there are several gradients that have been developed; however, we will concentrate on only one class of these, typically referred to as the morphological gradient.
Online access to SPIE eBooks is limited to subscribing institutions.