Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/8948

Title: Efficient representation, transmission and rendering of networked virtual environments on both desktop and mobile systems
Authors: JEHAES, Tom
Advisors: Lamotte, Wim
Van Reeth, Frank
Issue Date: 2008
Publisher: UHasselt Diepenbeek
Abstract: Over the last decade, Networked Virtual Environment (NVE) applications have greatly increased in popularity. The rst examples of these NVEs were mainly focussed on creating a virtual community in which users can navigate, communicate and interact. In recent years, however, gaming related features such as those found in role-playing games, racing games or rst person shooters have found their way into the world of networked virtual environments. Additionally, new technologies have been presented which have enabled the size of these virtual environments to grow exponentially resulting in massive online virtual communities or games. Related to the ever increasing size of virtual environments, huge amounts of data are often needed for representing their visual appearance. Contrary to earlier systems, where all the data could be stored locally, these new systems therefore need to resort to deploying large storage capacity servers from which users can request environment data at runtime. Ideally, transferring this data to the client needs to be done incrementally so the client can quickly provide an initial view of the environment. During application runtime, the environment data is continuously updated based on visibility calculations. In this work we will be targeting these large-scale dynamic networked virtual communities where application data needs to be requested at runtime. The main problem with these kinds of applications is the often disturbing scene buildup artefacts. The main research topic will therefore be focussed on reducing this scene buildup time while at the same time providing a high level of image quality coupled to a satisfactory user experience on both desktop and mobile systems. The research contributions presented in this thesis focus on three areas: data representation and streaming, objective Level-of-Detail selection and rendering on mobile devices. Firstly, we present a novel combined usage of imagebased and geometrical representations in dynamic networked virtual environments. Our first goal is to present the user with a su ciently realistic view of the virtual world as soon as possible while reducing disturbing scene buildup artifacts as well as guarantee interactive frame rates. We create image-based representations that are more compact to send than geometrical models so the client will sooner be able to render a more satisfactory view than if we would only be using incrementally transmitted models. The image-based models are also used in combination with traditional LoD techniques to speedup rendering while maintaining good display quality, which is our second goal. Run-time measurements of both client processing and network usage show the bene ts of using our system. Secondly, with regard to objective Level-of-Detail selection, the problem of rendering large virtual 3D environments at interactive framerates has traditionally been solved by using polygonal LoD selection techniques, which rely on a combination of metrics such as distance, projection size and importance. Our work on hybrid based data representations has shown that image-based techniques can be used to represent complex models in a compact way, thereby also decreasing the time needed for rendering. One of the questions that should be given some more consideration, however, is when to switch from polygonal rendering to image-based rendering in order to achieve a target framerate. We explore this topic further and provide a solution using an objective image quality metric which optimizes render quality. Several measurements, such as resulting framerate and image quality, are presented to show the bene t of our solution. Lastly, in order to allow users to access a large-scale online virtual environment at any time or place, some form of mobile access is needed using for instance a PDA or SmartPhone. We describe our approach for 3D rendering of a large virtual environment on PDA devices and discuss the main issues encountered. More speci cally, we use the results of our data representation and LoD selection research in order to limit the memory requirements and processing power needed. Furthermore, we investigate an alternative rendering scenario in which the client's view is created on a remote render server. We investigate the bene ts and drawbacks of this approach compared to the local rendering scenario.
URI: http://hdl.handle.net/1942/8948
Category: T1
Type: Theses and Dissertations
Appears in Collections: PhD theses
Research publications

Files in This Item:

Description SizeFormat
Final version21.86 MBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.