A Survey of Continuous Collision Detection

Abstract

Continuous collision detection (CCD) is a key technology in the field of virtual surgery, cloth simulation and robot motion planning. It can accurately detect the first time of contact between objects and returns collision information such as penetration depth, friction and repulsive force, etc., have a wide range of application and important research value. By analyzing the processing framework of continuous collision detection algorithm in detail, the current research status of continuous collision detection is systematically reviewed from perspectives of two phases respectively. In broad-phase, the recent achievements of space decomposition and sweep and prune are introduced. In narrow-phase, the research status of intelligent optimization based algorithm and image-space based algorithm is illustrated. Besides, the development of bounding volume hierarchy (BVH) is analyzed and discussed. After that, the performance and innovative achievements of self-collision detection in deformable objects are summarized and analyzed. Finally, the challenges and future trends of algorithm research are pointed out.