Measured data for paper submitted to EGRW'2000 , the paper named
BES: Best Efficiency Scheme

Building criteria for experiments:

BSP :BSP tree, maximum depth 16 (root of depth 0), maximum number of objects in leaves
were 2.

KD : KD tree. built with use of surface area heuristics. termination criteria the same as for
BSP tree

UG : uniform grid (uniform space subdivision) - Woos method of heteregeneous grids was
used, voxel density = 3.0

BVH : bounding volume hierarchy - the number of permutations to reform the hierarchy was 4.

AG : adaptive grids (Klimaszewski, Sedeberg), no subvoxel grids, f=1.5, m = 0.5, max objects
in voxel was 8.

RG : recursive grids, density 3.0, maximum depth of recursion was 3, max objects in voxel was 8.

HUG : hierarchy of uniform grids, nummber of levels was 3, connectivity was 0.0,
max objects in voxel was 8, heterogeneous grids were used.

O84 : octree with midpoint subdivision, maximum leaf depth was 6 (root node of depth 0),
maximum objects in leaves was 1.

O84A : Octree-R with surface area heuristics, the same criteria as for O84

O89 : octree with neighbour finding traversal algorithm, the same criteria as for O84.

O93 : octree using recursive traversal algorithm, the same criteria as for O84.

O93A : Octree-R, surface area heuristics, recursive traversal algorithm, the same criteria
as for O84.

Here are the data organized according to testing procedures:

The results for testing procedure TP_A are located
here (30 scenes, 10 parameters, 12 RASes)

The results for testing procedure TP_B are located
here (30 scenes, 10 parameters, 12 RASes)

The results for testing procedure TP_C are located
here (30 scenes, 10 parameters, 12 RASes)

The results for testing procedure TP_D are located
here (30 scenes, 10 parameters, 12 RASes)

Here are the data organized according to ray shooting acceleration schemes: