Automating the display of third person/stealth views of virtual environments

摘要

In order to gain a greater insight into the relationships that exist between entities in three-dimensional (3D) datasets, the scientific, engineering, and arts communities are increasingly using interactive visualization and virtual reality (VR) techniques. They have realized that inter-actively visualizing 3D datasets from different viewpoints makes it possible to achieve a better understanding of the underlying dataset structure. Viewpoints can be either static or dynamic as in an interactive fly-through. However, unskilled users often select flight paths (or viewing situations) that cause nauseous effects that detrimentally distract the user from the task at hand. Interactions between multiple users or virtual agents in a virtual environment present further challenges because it is necessary for the user to monitor multiple activities concurrently. If the user has to make decisions based on what is taking place in a complex virtual environment, then it is very important that correct and appropriate viewpoints are maintained. For example, flight simulator debriefing tools require first and third person viewing so that the actions that have taken place can be understood. In these situations there is a need to select multiple viewpoints for each participant. Consequently (because of the high cognitive load), maintaining control over a number of different viewpoints is very challenging. Within this paper the authors describe the real-time automatic display controller they have developed for third person/stealth views of a multi-participant virtual environment-where it is important for users to gain a global and localized understanding of the tasks being performed. They discuss important cinematic conventions and how these are set in the context of characterizing a 3D communication medium, as well as determining their role for optimal viewing parameters. The real time automatic display controller is of particular benefit to applications such as scientific visualization, flight simulation, engineering/architectural modeling, scene of accident reconstruction/analysis, and other complex human-system behavior analysis applications.