Thesis (M.Eng.)--Chulalongkorn University, 2006

An important problem in using three-dimensional data in computer graphics is effective data incompatibility, due to the various methods available to store three-dimensional data, such as wireframe modeling, winged-edge modeling, or surface modeling, etc. Each format uses different data structures, rendering them incompatible, and necessitating conversions from one format to another, which may result in incorrect or inconsistent data. The preferred structure for conversion output is usually the wireframe, because of its low complexity, and it is trivial to be converted from other models. However, it still lacks topological property of the three-dimensional data.This research proposes a method to convert three-dimensional data into a cellular structured space model, to efficiently store three-dimensional topological information. We use a wireframe as input, and then search for its points, edges, and possible faces for categorizing into cells with various dimensions. Finally, the cells are arranged into the appropriate cellular structure providing the topological relations between each component of the model.This research also explores possible applications using the proposed structure: we find the thinnest points in a solid, and we also smooth out a solid by converting its straight edges into curves. The experimental results show that the cellular structured space model correctly provides objects’ topological properties, and can be applied to correctly solving problems in computer graphics with reasonable time complexity.