Morphological Segmentation for Textures and Particles
Luc M. Vincent, Edward R. Dougherty · 2020
This chapter is concerned with image segmentation via the methods of morphological image processing. It is also concerned with regional texture-oriented segmentation and particle segmentation. The chapter focuses primarily on watershed segmentation, which has the advantages of being very general, usually accurate and fast, and applicable to both binary and gray-scale images. While granulometries can be employed to describe shape, more directly related to texture segmentation is their ability to measure changes in surface condition and in particle distribution processes, for instance electrophotographic images. In texture segmentation, an image is partitioned into regions, each of which is defined by a set of features characteristic to the microimage structure within it, this structure typically being viewed in terms of the small texture primitives composing it. As originally conceived for binary images, granulometries are employed to characterize size and shape information within granular binary images via the manner in which they are sieved through sieves of various sizes and shapes.