Hardware-accelerated collision detection for 3D virtual reality gaming

Abstract

Collision detection in simulation can easily become a bottleneck due to its computationally intensive nature. Recent developments in graphics hardware, however, offer a viable solution for rapid and efficient collision detection. The authors propose a new two-phase technique using the latest graphics hardware. In the broad phase, a scene graph is created to partition objects in a 3D environment for initial collision checking. In the narrow phase, a multiple-viewing volumes method is used to detect interferences between a convex model and a model of arbitrary geometry. First, the convex model is used to define six viewing volumes. The convex and arbitrary models are then rendered respectively within the defined viewing volumes. Finally, results of collision detection can be easily achieved by querying the occlusions between these rendered models in the image space. Compared with other collision detection algorithms, this method produces promising results and is successfully applied in our 3D virtual reality games.

Massively parallelized GA based optimal path planning for single and dual crane lifting in complex industrial environments

“…The LDIs are then used to analyze the repulsion forces on penetrating surfaces. Cai, et al [144] has proposed an image-space method using multiple projection directions to detect collision between convex objects.…”