An Intelligent 3D Navigation Interface for Large Virtual Environments

Abstract

Recent developments on graphics hardware and software have opened up new 3D interactive applications on personal computers. Most of these advances aim to provide better and faster graphics rendering. Recent researches suggest that motion-planning techniques can be successfully incorporated into the navigation control loop to improve the efficiency of 3D navigation in an architectural environment. However, a main limitation of this approach is on the scalability of the motion planner for large virtual environments. In this paper, we propose a novel approach to overcome this scalability problem. We limit the region of interest for path-finding to a window around the current viewpoint and incrementally update the roadmap in this window as the user viewpoint moves along. In order to facilitate the incremental update of the roadmap, we adopt a new data structure, called Rapidly-Exploring Random Tree (RRT), to reduce the run-time cost of building the connectivity roadmap. The incremental path planner has been implemented and incorporated into a Java3D-based VRML browser. We report experimental results of this improved user interface on a large-scale virtual environment. Computational bottlenecks of maintaining such an interface at run time are also analyzed. By extending the planning techniques to large-scale virtual scenes, we believe that this type of intelligent navigation will inspire better interactive 3D user-interface design in the future.