Year: 2006
Type: dissertation
Abstract: This thesis develops a strategy for polymerized volume data set construction.\nGiven a volume data set defined over a regular three-dimensional grid, a polymerized\nvolume data set (PVDS) can be defined as follows: edges between adjacent vertices of\nthe grid are labeled 1 (active) or 0 (inactive) to indicate the likelihood that an edge is\ncontained in (or spans the boundary of) a common underlying object, adding information\nnot in the original volume data set. This edge labeling “polymerizes” adjacent voxels\n(those sharing a common active edge) into connected components, facilitating\nsegmentation of embedded objects in the volume data set. Polymerization of the volume\ndata set also aids real-time data compression, geometric modeling of the embedded\nobjects, and their visualization.\nTo construct a polymerized volume data set, an adjacency class within the grid\nsystem is selected. Edges belonging to this adjacency class are labeled as interior,\nexterior, or boundary edges using discriminant functions whose functional forms are\nderived for three local adjacency classes. The discriminant function parameter values are\ndetermined by supervised learning. Training sets are derived from an initial\nsegmentation on a homogeneous sample of the volume data set, using an existing\nsegmentation method.\nThe strategy of constructing polymerized volume data sets is initially tested on\nsynthetic data sets which resemble neuronal volume data obtained by three-dimensional\nmicroscopy. The strategy is then illustrated on volume data sets of mouse brain\nmicrostructure at a neuronal level of detail. Visualization and validation of the resulting\nPVDS is shown in both cases. Finally the procedures of polymerized volume data set construction are\ngeneralized to apply to any Bravais lattice over the regular 3D orthogonal grid. Further\ndevelopment of this latter topic is left to future work.
Author: Prathyusha Aragonda
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2901453720