| Popis: |
Navigation is one of the most fundamental tasks in virtual environments, that requires timely collision detection and collision avoidance mechanisms. It was shown that when navigation is constrained to moving on a terrain, the problem of collision detection can be reduced from 3D to 2D case, by encoding all 3D objects that constitute travel obstacles into a terrain elevation image map, using a dedicated color. The original approach was developed for static environments explored by a single user. The new improved system, presented in this article, is capable of processing collisions for multiple travelers, including autonomous agents, in dynamically changing environments. Implementation details of the new system are presented, with a number of case studies and a discussion of future work. |
