Invisible space skipping with adaptive granularity for texture-based volume rendering
View Patent ↗A method of accelerating texture-based volume rendering is provided. The method includes the following steps. A plurality of levels of granularity comprising a plurality of sub-volumes organized in a hierarchical structure are built. The hierarchical structure is ordered top-down from a coarsest level of granularity to a finest level of a granularity. A visibility factor is determined for each of the plurality of sub-volumes. The visibility factor represents a ratio of visible voxels in the corresponding sub-volume. The plurality of sub-volumes are traversed top-down from the coarsest level of granularity to the finest level of granularity. While traversing each of the plurality of sub-volumes, the corresponding sub-volume is rendered and all children of the rendered sub-volume are skipped, if the visibility factor is greater than given threshold.
1. A method of accelerating texture-based volume rendering, comprising:
(a) building a plurality of levels of granularity comprising a plurality of sub-volumes organized in a hierarchical structure, wherein the hierarchical structure is ordered top-down from a coarsest level of granularity to a finest level of a granularity;
(b) determining a visibility factor for each of the plurality of sub-volumes, wherein the visibility factor represents a ratio of visible voxels in the corresponding sub-volume;
(c) traversing the plurality of sub-volumes top-down from the coarsest level of granularity to the finest level of granularity; and
(d) while traversing each of the plurality of sub-volumes in step (c), rendering via a processor the corresponding sub-volume and skipping all children of the rendered sub-volume, if the visibility factor is greater than a given threshold.
2. The method of claim 1 , further comprising:
if a volume comprising the plurality of sub-volumes is bricked, loading the plurality of sub-volumes from a main memory to a graphics memory with adaptive granularity,
wherein the step loading is performed if the corresponding visibility factor is greater than a loading threshold; and
wherein the step of loading is performed with the same or coarser level of granularity as step (d).
3. The method of claim 2 , wherein the step of loading the plurality of sub-volumes, comprises:
overlapping neighboring bricks to ensure proper interpolation during rendering.
4. The method of claim 3 , wherein the step of loading the plurality of sub-volumes, further comprises:
offsetting a sampling position by a half voxel to reduce the overlapping by one voxel.
5. The method of claim 3 , wherein the step of loading the plurality of sub-volumes, further comprises:
if the rendering at step (d) is unshaded, expanding the brick by one voxel on a high-end with no expansion on a low end; and
if the rendering at step (d) is shaded and computes gradients on-the-fly, expanding the brick by two voxels on the high-end and one voxel on the low end.
6. The method of claim 1 , wherein the visibility factor represents a percentage of visibly voxels in the corresponding sub-volume.
7. The method of claim 1 , wherein the step of determining a visibility factor for each of the plurality of sub-volumes, comprises:
combining a plurality of properties associated with the corresponding sub-volume to determine the visibility factor.
8. The method of claim 1 , wherein the step of determining a visibility factor for each of the plurality of sub-volumes, comprises:
computing the visibility factor for a first portion of the plurality of sub-volumes;
propagating respective property values of said parent and child sub-volumes for a remaining portion of the plurality of sub-volumes; and
during the step of propagating, updating the visibility factor for the corresponding sub-volume.
9. The method of claim 1 , wherein the step of traversing the plurality of sub-volumes, comprises:
utilizing a common iterator and a plurality of replaceable filters for efficient selection of sub-volumes in traversal.
10. The method of claim 9 , further comprising:
adding new filters with new sub-volume selection criteria without changing the common integrator and the plurality of replaceable filters.
11. The method of claim 9 , wherein the common integrator and the plurality of replaceable filters are implemented with an inline method.
12. The method of claim 9 , wherein the common integrator and the plurality of replaceable filters are implemented with C++ templates.
13. The method of claim 1 , wherein the step of determining a visibility factor, comprises:
determining a visibility factor with adaptive granularity.
14. A program storage device having instructions stored thereon for accelerating texture-based volume rendering, comprising:
(a) building a plurality of levels of granularity comprising a plurality of sub-volumes organized in a hierarchical structure, wherein the hierarchical structure is ordered top-down from a coarsest level of granularity to a finest level of a granularity;
(b) determining a visibility factor for each of the plurality of sub-volumes, wherein the visibility factor represents a ratio of visible voxels in the corresponding sub-volume;
(c) traversing the plurality of sub-volumes top-down from the coarsest level of granularity to the smoothest level of granularity; and
(d) while traversing each of the plurality of sub-volumes in step (c), rendering the corresponding sub-volume and skipping all children of the rendered sub-volume, if the visibility factor is greater than a given threshold.