Layered hypervideo document database system: Design and modeling of hypervideo document database

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering and Computer Science


P. Bruce Berra


Video coding, Petri net, Hypervideo, Document database

Subject Categories

Computer Sciences | Physical Sciences and Mathematics


The study of hypervideo document (HVD) database systems is a rather new area of research. These documents place additional requirements on data models due to the characteristics of hypervideo documents. These characteristics include employment of active user interactions, dynamic temporal structure of the document, and large data. To satisfy these requirements, we designed the Layered Hypervideo Document Database System (LHDDS) what is based on the new data model called the Layered Hypervideo Document Data Model (LHDM). The LHDDS is a, general and open database architecture that can support hypervideo document applications as well as other conventional multimedia document applications. Each of its four layers, physical, logical, composite, and application, encapsulates a different level of data abstraction. The physical layer (PL) encapsulates hypervideo documents at the binary data stream. The logical layer (LL) encapsulates hypervideo documents as sets of consecutive frames. The composite layer (CL) focuses on the dynamic nature of interactive document structures. The application layer (AL) encapsulates hypervideo documents within the end-user level of abstraction by supporting applications such as authoring tools, video kiosks, and video-intensive WWW applications. The LHDM and LHDDS provide data encapsulation, openness to future technology, data sharing/reusing, and simplification of problems related to hypervideo document database systems. The contributions of each layer varies. The PL has fast random access of frames supported for MPEG video data; the LL has a storage efficient indexing structure based on hypervideo objects to support fast and flexible access to various units of an HVD; the CL has an interactive temporal model for interactive multimedia documents to support the dynamic structure of hypervideo documents; and the AL has convenient and data-safe development of applications support through interfaces of each layer. In the composite layer, a new interactive temporal multimedia document structure model, called a Transition Object Composition Petri-Net (TOCPN), is provided with a structure-addressing query language and query processing algorithms based on the structure of the TOCPN model. With the new composition model and LHDM, LHDDS can support functionality such as active user interaction, query formulation/modification, content-based retrieval, dynamic adaptation for QoS, synchronization management, and dynamic structural navigation.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.