Acceleration factor w.r.t. Rapid:

SGI Onyx2 IR

DEC Alpha 4D60

Pentium (3Dfx)

Pentium (soft)

1

2

3

4

5

1

50

100

150

200

Rigid Object

Deformable Object

Diapositive précédente

Diapositive suivante

Revenir à la première diapositive

Afficher la version graphique

Here we give the mean acceleration factor of our method with respect to the OBB-tree method implemented in RAPID.

The left graph concerns a rigid object. This is the RAPID best case, because we neglect the pre-computation time. We almost have an acceleration factor of 5 with our method on OpenGL graphics hardware. On a personal PC (we took a pentium 2 333MHz) with or without a gaming Voodoo board, our algorithm is only two times slower than RAPID which is of complex programming.

We also see that if we take into account the RAPID pre-computation times, as is the case in the right graph, our method is more than one hundred fifty times faster than OBB-tree for our collision detection problem. On the Onyx we reach an acceleration factor of one hundred and ninety .

On the PC, our algorithm replies fifteen times faster than RAPID, even with no graphic specific hardware.