The design of a dynamic Voronoi Map Object (VMO) model for sustainable forestry data management

The Voronoi diagram is a powerful geometric structure attractive to many applications, especially when it is developed with computers. This thesis investigates the flexibility of such a geometric structure, called the dynamic Voronoi diagram of points and line segments, applied to GIS, geographical information systems. In particular, the question concerned in the thesis is: Given forestry as an application domain, is the dynamic Voronoi diagram a useful GIS data model to support applications concerned with the sustainability of forestry?; The thesis provides a strong positive answer to the above question. By reviewing the characteristics of sustainable forestry, the thesis summarizes the corresponding requirements for the supporting spatial data models. After examining traditional spatial data models and data structures used in current GIS systems, the thesis argues that the dynamic Voronoi data model can support integration of the topology and geometry and satisfies all the requirements for a dynamic GIS. The argument is supported by a computer implementation of the dynamic Voronoi GIS in its primitive form and by presenting common GIS operations over the Voronoi data model.; The thesis contributes to the development of a dynamic Voronoi GIS by proposing a Voronoi Map Object (VMO) model which removes the limitation of memory occupation for large Voronoi diagrams. The VMO model is achieved by partitioning the Voronoi diagram into subdiagrams and by representing them with a hierarchical object structure. Each node on the structure is a VMO and support full topology and geometry about, and operations on the object. The thesis describes the algorithm for partitioning (and pasting) a Voronoi diagram and the formalism of the VMO model.; Finally, the thesis discusses design issues for a forestry data management system using the VMO model. The discussion covers the object model, the dynamic model, the functional model, and the software architecture of the system. The applications of the VMO model for parallel processing of spatial problems and for automated map generalization are also briefly discussed.