Research on Collision Detection Algorithm Based on AABB
Wang Xiao-rong, Wang Meng, Chungui Li · 2009
An improved collision detection algorithm based on AABB is presented. During the global search, each axis is cut into a series of segments containing the same number of AABBs' projection intervals, and Shell sort is adopted to sort projection lists, not insertion sort. This will avoid needless intersecting test of AABB. During the local detection, the amount of byte of AABB bounding-volume for internal node is reduced according to the constructing process of AABB tree, and leaf nodes are wiped from tree structure The storage of AABB tree is compressed. This method can save a large amount of space and speed up the algorithm. Experiments indicate that the improved algorithm reduce detection time for the same models.