Aachen : Shaker, 2002
Berichte aus der Informatik
Darmstadt, TU, Diss., 2001
Today the Internet is a global communication and business platform. Its breakthrough began with the hypertext transfer protocol (HTTP), which sits on top of the Internet protocol TCP/IP and allows the hypertext based interconnection of documents through links in conjunction with a corresponding markup language (HTML). It provides features for the inline visualization of various media types such as image, video and sound. But despite this rich hypermedia functionality, state of the art applications do not exploit the benefits of interactive 3D graphics. It is well known, that: »An image tells more than a thousand words.« (Chinese saying). In many applications, replacing a static image by an interactive 3D application could significantly raise the value to the users by allowing them to actively experience the natural qualities of a model by performing actions and interacting with the model. The advantages of active experience are cherished for long: »I hear and I forget. I see and
I remember. I do and I understand.« (Konfuzius). Hypermedia applications could benefit from interactive dynamic 3D visualization.
This work shows how. It analyzes the state of the art technologies, unveils major problems in the production process, introduces a component based concept and a respective example implementation. Dynamic, i.e. animated 3D models are the heart of a 3D application. Therefore, first the traditional ways of creating 3D models and animations - the modeling and animation pipelines - are reviewed. It is specified, which requirements must be met by an animated 3D model such that it can be used in an interactive hypermedia application. These requirements must be reflected in the 3D hypermedia application description format. The state of the art technologies are analyzed and the current application description format which comes closest to what would be needed is determined. Based on this, a component based concept for the description of 3D hypermedia applications has been developed, which is independent of the file format. The 3D model component and the animation, i.e. the functional
component, which are the heart of the component based concept are introduced first. Second this work analyzes 3D navigation. Intuitive navigation is a key feature required in 3D worlds. The most frequently quoted reason why people are reluctant to use 3D applications is a loss of orientation in the virtual world, caused by bad navigation support. Object-centered navigation is a concept of intuitive motion within a virtual world. Its key aspect is to specify motion steps relative to the object(s) that is (are) in focus. Objectcentered viewpoints provide a discrete motion space and allow for an easy navigation. Tests have shown that the method is preferred by nonexpert users when compared to standard navigation interfaces. Third, the state of the art hypermedia technology with focus on its limited 3D integration capacities is reviewed. Based on the previously specified requirements, a model for fully integrating component based 3D applications into a hypermedia environment is suggested
and validated in an implementation. Application examples are presented. The first one shows an interactive three dimensional product catalogue, where each part of the animated 3D product model is interconnected with textual and visual media and vice versa. The product parts can be »touched« and moved. Actions in the model are reflected in the related hypertext document and vice versa. The second example is based on the 3D model of a large building, a construction plant. Both illustrate fully fledged 3d hypermedia applications.
The key results of this work show: first of all a concept for effectively integrating 3d models into a hypermedia application context, second that the suggested concepts allows for significant improvement of application development efficiency in terms of time savings, cost reduction and reusability, third that object-centered navigation is a preferred concept for navigating virtual worlds in an intuitive fashion, fourth that these concepts can be exploited in »real world« applications and fifth an exemplary reference implementation. Future work would be e.g. advancements of the object-centered navigation by extending the here presented simple method to a generic path finding algorithm.