Volumetric feature extraction of 3D images defined over hexagonal prism lattice

Abstract

A 2D honeycomb image is a hexagonal array of what are called pixels and the pixel values represent intensity or color information of the corresponding digital image at various positions addressed by zig-zag and column numbers. A 2D image could also be informally called as image honeycomb or an arrangement of pixel values in a hexagonal array. A 3D hexagonal prism lattice image is viewed as an ordered sequence of 2D image honeycombs arranged in the z-direction and the 3D arrangement of voxel values is called a hexagonal prism lattice of voxel values. Most of the 3D hexagonal prism lattice image processing operations are similar to those of 2D image honeycomb processing. 3D images are processed with the help of 3D scanning windows, whereas 2D images are processed with the help of 2D scanning windows. A digital image defined over a hexagonal prism lattice has better curvilinear properties when compared to an image defined over a rectangular prism lattice. Processing of hexagonal lattice based images cannot be carried out using simple rectangular 2D or 3D scanning mask. This paper describes a special type of hexagonal lattice based scanning masks for processing such images. As a special case study, edge detection and skeletonization of 3D hexagonal prism lattice images are presented.