Dynamic radial view based culling for continuous self-collision detection

Abstract

Figure 1: (a) (b) The clusters are fixed in the radial view based culling (RVBC) method at runtime. (c) (d) Our method dynamically merges the atomic clusters to improve the overall performance of continuous self-collision detection over RVBC. The speedup of our method compared to RVBC is 5.2× in this octopus example. AbstractThe radial view-based culling (RVBC) method has been presented for continuous self-collision detection to efficiently cull away noncolliding regions. While this technique mainly relies on the segmented clusters of the reference pose and the associated fixed observer points, it has several drawbacks during the animation and the reduced cost of executing collision detection is limited. We thus present a modified framework to improve the culling efficiency of RVBC. At the preprocessing stage, we segment the closed deformable mesh according to not only the attached skeleton but also the triangle orientations, in order to minimize the collision checks of triangles in a cluster. At the runtime stage, we dynamically merge adjacent clusters and update the positions of observer points if the merged shape is nearly convex. This strategy minimizes the number of triangles in different clusters that required collision check. Our framework can be easily integrated with bounding volume hierarchies to boost the culling efficiency. Experimental results show that our framework achieves up to 5.2 times speedup over the original RVBC method and even more times over the recent techniques.

A multi‐GPU finite element computation and hybrid collision handling process framework for brain deformation simulation

“…However, for the triangles in the same BV, the BVH traversal is required to perform at the lowest level, even though the entire mesh is collision free. A radial view‐based culling method has been presented for continuous self‐collision detection to efficiently cull away noncolliding regions. This method requires the mesh to be attached with a skeleton.…”

“…BVH: A BVH is a tree structure on a set of geometric objects that accelerate queries. BVHs have been widely used for collision detection, in which different BVs are selected . Recently, there has been considerable interest in developing parallel techniques to construct or update the BV hierarchies for deformable models.…”

Interactive continuous collision detection for topology changing models using dynamic clustering

“…However, these techniques, which include an expensive precomputation step, may therefore not work well on models undergoing topological changes. Recently, [Wong et al 2013; Wong et al 2014] proposed a technique that accelerates continuous collision detection by performing radial view-based culling based on the skeleton structure. However, the skeleton needs to be precomputed and the overhead for models undergoing topology changes can be high.…”

“…These clusters have been used to compute tight-fitting bounding volume hierarchies from the bottom up [Tan et al 1999] or used directly for collision detection [Wong et al 2013; Wong et al 2014; Ehmann and Lin 2001; Schvartzman et al 2010; Wong and Baciu 2014]. However, all these techniques either perform some precomputation or assume fixed-mesh connectivity during the decomposition step.…”

“…Note that we need to ensure that each $\mathrm{scriptLC}$ has no self-collisions, and that we want to decompose the boundary of a mesh into as few $\mathrm{scriptLC}$s as possible. Our formulation uses the notion of an observation point , introduced by [Wong et al 2013; Wong et al 2014], to compute these clusters using an adaptive scheme.…”