Research On Collision Detection Algorithm Based On Intersection Progressive Of Bounding Box

Collision detection is the behavior to judge whether the relationship between the position of moving objects in the virtual scene is conflict. It takes an important place in the field of virtual reality, robot path planning, computer graphics and simulation etc. At present domestic and international research on the collision detection problem has a lot of valuable research results. However, with the big scene in virtual reality technology is widely used in the new field. People’s requirement for the authenticity of virtual environment and interaction and real-time for user improves constantly, improving the speed of collision detection effectively has become an urgent task at present. Aiming at this problem, this thesis improves the algorithm from the overall positioning the pairs of collision objects in the scene and the structure of the detection structure for the reduction of the burden on collision positioning calculation in the big scene. It also effectively improves the efficiency of the collision detection between the complex structure model and complex scenes. It achieves the requirements of real-time detection.In this thesis, it improves the efficiency of collision detection algorithm based on bounding box along the axis from two aspects of global detection and local detection. First of all, in the whole detection process, due to correlation of space and time of the mobile object, this thesis takes the delivery location algorithm. This algorithm takes method of dividing partition detection range based on the object and of the dimension reduction to sort the projection interval. This method can effectively avoid error interference of obviously disjoint of the object in the global detection and enhance the efficiency of the global localization algorithm. In the local detection process, this thesis steps close to the intersection progressively and to determine whether the intersection of two geometric model from the macro to the micro. In the progressive process, the algorithm regards the intersection part of the two AABB bounding box as a new subspace. In this subspace, the following intersection test for models in this space repeat the above steps. The method can reduce the structural of redundant bounding box in storage space and save storage space and construction time. It enhances the efficiency of collision detection fundamentally.Complex intersection test and application of the geometric model in this thesis is based on the BIM technology platform Revit API. The experimental results show that, the algorithm can quickly and effectively solve the problem of efficiency of collision detection in the virtual space in the global and local detection. The test verifies that this algorithm is of great help in the efficiency of the running speed of large-scale projects in utilization rate of time and space. It also has the practical significance and theoretical value in the research of collision detection.