Real-Time Continuous Collision Detection Based on Swept Volume and Depth Texture

We present a novel algorithm to perform continuous collision detection in distributed virtual environment. Our algorithm consists of four stages: (1) generating depth texture of model; (2) computing depth texture of swept volume by model depth texture; (3) mapping depth image of swept volume on a face of bounding box; (4) comparing depth of swept volume with depth of another object to detect collisions. We have implemented the algorithm and the results show that our algorithm is able to perform fast and accurate continuous collision detection and have fixed executing time to suit for real-time applications