Ivannikov Institute for System Programming of the RAS


Combined strategy for efficient collision detection in 4D planning applications.

Authors

Vitaly Semenov & Konstantine Kazakov & Vladislav Zolotov & Huw Jones & Stephan Jones

Abstract

Collision detection is the most time consuming component of many geometric reasoning applications like CAD/CAM, robotics and automation, computer graphics, and virtual reality. Performance is also a crucial factor for 4D planning applications targeted on scheduling of large industry projects by means of identifying and resolving of clashes in complex dynamic scenes, containing thousands and millions of objects. Many methods have been proposed and successfully implemented to deal with different statements and peculiarities of the dynamic collision detection problem. In this paper, we develop and analyze a combined strategy that exploits spatial and temporal coherence of scenes appearing in 4D planning applications. The strategy takes into account complexity and large scales of scenes, evolving character of fragments, significant temporal latency, and availability of identical objects. The strategy combines spatial decomposition, bounding volumes and temporal analysis methods yielding substantially faster collision detection than previous known methods.

Text of article

Edition

Computing in Civil and Building Engineering, Proceedings of the International Conference, 30 June-2 July, Nottingham, UK. Publisher: Nottingham University Press, 2010. Pp. 31 (full paper 6 p.p. published on CD).

Research Group

System integration and multi-disciplinary collaborative environments

All publications during 2010 All publications