Simple and rapid collision detection using multiple viewing volumes

Collision detection is required by most virtual reality (VR) applications, and in recent years, collision detection algorithms utilizing graphics hardware capabilities have been attracting the interests of the VR community. In this paper, we present a novel approach to detecting interferences between a convex model and a model of arbitrary geometry with commercially available graphics hardware. The approach first defines six viewing volumes through the bounding box of the convex model, then renders the two models within the defined viewing volumes respectively. The results of collision detection can simply be obtained by querying occlusions between the rendered models in the image space. The approach has been implemented and several scenarios have been tested. Experimental results have been given and compared with other typical collision detection algorithms.

• Type: Conference
• Year: 2004
• Where: VRCAI
• Authors: Zhaowei Fan, Huagen Wan, Shuming Gao
• Keywords: Collision Detection, Collision Detection Algorithms, Convex Model, Virtual Reality, VRCAI 2004