Kinetic bounding volume hierarchies for deformable objects

Abstract

We present novel algorithms for updating bounding volume hierarchies of objects undergoing arbitrary deformations. Therefore, we introduce two new data structures, the kinetic AABB tree and the kinetic BoxTree.The event-based approach of the kinetic data structures framework enables us to show that our algorithms are optimal in the number of updates. Moreover, we show a lower bound for the total number of BV updates, which is independent of the number of frames.We used our kinetic bounding volume hierarchies for collision detection and performed a comparison with the classical bottomup update method. The results show that our algorithms perform up to ten times faster in practically relevant scenarios.

Balanced Hierarchies for Collision Detection between Fracturing Objects

“…Some of the popular techniques for collision detection between deforming objects, such as visibility-based culling [11], distance fields [9,29], layered depth images [13], or spatial hashing [30], assume no pre-processing of the surfaces. Bounding volume hierarchies (BVHs) have also proved to be competitive when updated efficiently, for complete collision detection between deforming objects [32,7,18,15,19,34], or as a first culling stage in combination with visibility-based algorithms [12].…”

“…We consider two different types of updates: (i) refitting of BVs, and (ii) restructuring of the tree. Efficient refitting has been extensively studied, and it includes approaches such as hybrid top-down and bottom-up updates [18], lazy update with enlarged BVs [21], fully top-down update for reduced deformable models [15], or eventbased refitting in the framework of kinetic data structures [34]. Efficient restructuring of the trees, however, is a less explored problem.…”

“…One conclusion that can be drawn from this is that our dynamic restructuring algorithm can be complemented with existing work for efficiently refitting BVs or rebuilding parts of the BVH [34,19]. Therefore, we decided to fully rebuild the BVH of the apple whenever the fitting quality Q > 2.3.…”

“…We show that, in cases where objects fracture progressively, dynamically updating BVHs with our approach is up to 20 times faster than fully reconstructing them from scratch. Moreover, complemented with existing efficient techniques for refitting BVHs [19,34], our dynamic restructuring algorithms can provide these update speedups with almost no penalty on the performance of collision detection.…”

“…We build our approach on bounding volume hierarchies (BVHs), which have proved to be a successful acceleration data structure for collision detection between rigid bodies [14,10,16,8], collision detection between deformable bodies [32,7,18,15,19,31,34], and also as a first step of culling in combination with other methods [12].…”