Abstract
Collision detection is mainly to judge whether there is intersection between virtual models, which means there is collision. Bounding box is one of the important methods of collision detection. It uses regular geometry with simple structure to replace the complex model to be detected. A simple point cloud model usually contains hundreds or thousands of polygons when meshed. When collision tests are performed directly on the geometry of two model objects, the calculation process is relatively complex. In order to reduce the computational cost, the bounding box test of the model was performed before the geometry intersection test. Only when the bounding box has a collision, can further accurate intersection detection be carried out. This paper mainly introduces the content and significance of the bounding box, and compares four kinds of common bounding boxes and their advantages and disadvantages as well as the application scenarios of the bounding box. Firstly, the content and significance of the bounding box are expounded. Secondly, it is analyzed according to different bounding boxes. Finally, the different bounding box algorithms are summarized, and the advantages and disadvantages of different bounding boxes and application scenarios are pointed out.