Construction of bounding volume hierarchies with SAH cost approximation on temporary subtrees

Abstract Advances in research on quality metrics for bounding volume hierarchies (BVHs) have shown that greedy top-down SAH builders construct BVHs with superior traversal performance despite the fact that the resulting SAH values are higher than those created by more sophisticated builders. Motivated by this observation we examine three construction algorithms that use recursive SAH values of temporarily constructed SAH-built BVHs to guide the construction and perform an extensive evaluation. The resulting BVHs achieve up to 37% better trace performance for primary rays and up to 32% better trace performance for secondary diffuse rays compared to standard plane sweeping without applying spatial splits. Allowing spatial splits still achieves up to 18% resp. 15% better performance. While our approach is not suitable for real-time BVH construction, we show that the proposed algorithm has subquadratic computational complexity in the number of primitives, which renders it usable in practical applications. Additionally, we describe an approach for improving the forecasting abilities of the SAH-EPO ray tracing performance predictor which also increases its relevance for primary rays.

[1]  Dinesh Manocha,et al.  Fast BVH Construction on GPUs , 2009, Comput. Graph. Forum.

[2]  Timo Aila,et al.  On quality metrics of bounding volume hierarchies , 2013, HPG '13.

[3]  Philipp Slusallek,et al.  Object partitioning considered harmful: space subdivision for BVHs , 2009, High Performance Graphics.

[4]  Shiqi Li,et al.  Fast BVH construction and refit for ray tracing of dynamic scenes , 2013, Multimedia Tools and Applications.

[5]  John Salmon,et al.  Automatic Creation of Object Hierarchies for Ray Tracing , 1987, IEEE Computer Graphics and Applications.

[6]  I. Wald,et al.  On fast Construction of SAH-based Bounding Volume Hierarchies , 2007, 2007 IEEE Symposium on Interactive Ray Tracing.

[7]  Ingo Wald,et al.  Ray tracing deformable scenes using dynamic bounding volume hierarchies , 2007, TOGS.

[8]  Andreas Dietrich,et al.  Spatial splits in bounding volume hierarchies , 2009, High Performance Graphics.

[9]  S. Popov,et al.  Experiences with Streaming Construction of SAH KD-Trees , 2006, 2006 IEEE Symposium on Interactive Ray Tracing.

[10]  B. Walter,et al.  Fast agglomerative clustering for rendering , 2008, 2008 IEEE Symposium on Interactive Ray Tracing.

[11]  John Dingliana,et al.  A Cost Metric for Scene-Interior Ray Origins , 2009, Eurographics.

[12]  Michael C. Doggett,et al.  SAH guided spatial split partitioning for fast BVH construction , 2016, Comput. Graph. Forum.

[13]  Vlastimil Havran,et al.  Heuristic ray shooting algorithms , 2000 .

[14]  Timo Aila,et al.  Fast parallel construction of high-quality bounding volume hierarchies , 2013, HPG '13.

[15]  Kellogg S. Booth,et al.  Heuristics for ray tracing using space subdivision , 1990, The Visual Computer.

[16]  Timo Aila,et al.  Understanding the efficiency of ray traversal on GPUs , 2009, High Performance Graphics.

[17]  Michael Goesele,et al.  Recursive SAH-based Bounding Volume Hierarchy Construction , 2016, Graphics Interface.

[18]  Guy E. Blelloch,et al.  Efficient BVH construction via approximate agglomerative clustering , 2013, HPG '13.

[19]  T. Akenine-Möller,et al.  Bonsai: Rapid Bounding Volume Hierarchy Generation using Mini Trees , 2015 .

[20]  Vlastimil Havran,et al.  Incremental BVH construction for ray tracing , 2015, Comput. Graph..

[21]  I. Wald,et al.  On building fast kd-Trees for Ray Tracing, and on doing that in O(N log N) , 2006, 2006 IEEE Symposium on Interactive Ray Tracing.

[22]  Kirill Garanzha,et al.  Simpler and faster HLBVH with work queues , 2011, HPG '11.

[23]  Jacopo Pantaleoni,et al.  HLBVH: hierarchical LBVH construction for real-time ray tracing of dynamic geometry , 2010, HPG '10.