Research on path planning of intelligent virtual human in distributed virtual environment

Virtual human with intelligent behavior plays an important role in distributed virtual environment. The paper studies path planning that is an essential intelligent behavior of virtual human. According to the features and requirements of distributed virtual environment, we propose a method which combines global static planning and local dynamic planning. Global static planning method builds `reachable point´ based on the abstract modeling of virtual environment, then adopts A* or Dijkstra algorithms to find optimal path and builds a matrix in advance. Local dynamic path planning takes advantage of `virtual sensing´ of the virtual human to realize obstacle-avoiding path adjustment in dynamic environment. Finally, we present the approach that implements the walking animation of Web-based virtual human using VRML & H-Anim standards. We applied our approach to implement a virtual guide in a Web-based virtual environment and got satisfying results.