The Application Research On Hybrid Collision Detection Algorithm In Virtual Reality

In computer graphics, robot path planning, virtual assembly, computer animation and other fields, the problem of Collision Detection has a long research history. Collision Detection has become a hotspot. In the virtual environment, not only the reality and immersive need the height of model simulation, but also the precise collision detection between objects and the effect after the collision of real and real-time simulation.In this paper, sphere bounding box algorithm, OBB bounding box and the random collision detection algorithm based on improved particle swarm optimization algorithm are studied. In order to improve real-time of collision detection algorithm, the intelligent optimization algorithm is introduced. A complex three-dimensional model of the collision problem is transformed into a simple two-dimensional discrete space optimization problem. In random collision detection algorithm, in order to reduce interference miss some elements and to improve the accuracy of collision detection,the basic geometric elements of the OBB bounding box instead of the basic geometric elements as the feature samples.By studying a variety of collision detection algorithm, Thesis presents a hybrid collision detection algorithm, which is suitable for multi-body collision detection system. First, in the virtual scene, algorithm creates sphere bounding box and OBB hierarchy bounding box for each model, then some models are excluded by the testing of the intersection of sphere bounding boxes. Using the OBB bounding box intersection test excludes the non-intersect part of model between the possible intersection two models. This algorithm can quickly excludes the non-intersect models, and avoid the large search space due to target slow or premature, and reduce the hierarchical bounding algorithm takes a lot of memory space and slow updates.Finally, it is authentication that hybrid collision detection algorithm is efficiency by the experiment, and compare to the classic OBB collision detection algorithms and the random collision detection algorithm based on improved particle swarm algorithm, the algorithm is real-time and efficient.