System, method, and computer program product for pre-filtered anti-aliasing with deferred shading

Patent metadata

Field	Value
Publication number	US-9547931-B2
Application number	US-201514704875-A
Country	US
Kind code	B2
Filing date	May 5, 2015
Priority date	May 5, 2014
Publication date	Jan 17, 2017
Grant date	Jan 17, 2017

Abstract

A system, method, and computer program product are provided for generating anti-aliased images. The method includes the steps of assigning one or more samples to a plurality of clusters, each cluster in the plurality of clusters corresponding to an aggregate stored in an aggregate geometry buffer, where each of the one or more samples is covered by a visible fragment and rasterizing three-dimensional geometry to generate material parameters for each sample of the one or more samples. For each cluster in the plurality of clusters, the material parameters for each sample assigned to the cluster are combined to produce the aggregate. The combined material parameters for each cluster are stored in an aggregate geometry buffer. An anti-aliased image may then be generated by shading the combined material parameters.

First claim

What is claimed is:

1. A method comprising:
assigning one or more samples to a plurality of clusters, each cluster in the plurality of clusters corresponding to an aggregate stored in an aggregate geometry buffer, wherein each of the one or more samples is covered by a visible fragment;
rasterizing three-dimensional geometry to generate material parameters for each sample of the one or more samples;
for each cluster in the plurality of clusters, combining the material parameters for each sample assigned to the cluster to produce the aggregate; and
storing the combined material parameters for each cluster in the aggregate geometry buffer.

2. The method of claim 1 , further comprising, prior to assigning the one or more samples to the plurality of clusters, computing a mean position and a mean normal vector for the one or more samples in a screen-space region.

3. The method of claim 2 , further comprising defining a first cluster of the clusters as including a first sample of the one or more samples that is farthest from the mean position.

4. The method of claim 3 , further comprising defining a second cluster of the clusters as including a second sample of the one or more samples that is farthest from the first sample.

5. The method of claim 4 , further comprising:
determining that the remaining samples of the one or more samples are within a distance t of either the first cluster or the second cluster; and
assigning each of the remaining samples to the nearest of the first cluster and the second cluster.

6. The method of claim 1 , wherein the aggregate represents a statistical distribution of sub-pixel geometry within a screen-space region.

7. The method of claim 6 , wherein the screen-space region is a pixel.

8. The method of claim 1 , further comprising shading the combined material parameters to produce an anti-aliased image.

9. The method of claim 1 , wherein the assigning of the one or more samples to the plurality clusters is based on depth values of the one or more samples.

10. The method of claim 1 , wherein the assigning of samples to clusters is based on positions of the visible fragment that covers each sample.

11. The method of claim 1 , wherein the assigning of the one or more samples to the plurality clusters is based on normal vector data for the visible fragment that covers each sample.

12. The method of claim 1 , wherein a pixel includes n samples and c clusters are defined for the pixel, where c is less than n.

13. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform steps comprising
assigning one or more samples to a plurality of clusters, each cluster in the plurality of clusters corresponding to an aggregate stored in an aggregate geometry buffer, wherein each of the one or more samples is covered by a visible fragment;
rasterizing three-dimensional geometry to generate material parameters for each sample of the one or more samples;
for each cluster in the plurality of clusters, combining the material parameters for each sample assigned to the cluster to produce the aggregate; and
storing the combined material parameters for each cluster in the aggregate geometry buffer.

14. A system comprising:
a memory configured to store an aggregate geometry buffer; and
a plurality of multithreaded processing units that are included within a parallel processor and are coupled to the memory and configured to:
assign one or more samples to a plurality of clusters, each cluster in the plurality of clusters corresponding to an aggregate stored in the aggregate geometry buffer, wherein each of the one or more samples is covered by a visible fragment;
rasterize three-dimensional geometry to generate material parameters for each sample of the one or more samples;
for each cluster in the plurality of clusters, combine the material parameters for each sample assigned to the cluster to produce the aggregate; and
store the combined material parameters for each cluster in the aggregate geometry buffer.

15. The system of claim 14 , wherein, prior to assigning the visible samples to the clusters, the plurality of processing cores are further configured to compute mean position and mean normal vector data for the one or more samples in a screen-space region.

16. The system of claim 15 , wherein the plurality of processing cores are further configured to define a first cluster of the clusters as including a first sample of the one or more samples that is farthest from the mean position.

17. The system of claim 14 , wherein the aggregate represents a statistical distribution of sub-pixel geometry within a screen-space region.

18. The system of claim 17 , wherein the screen-space region is a pixel.

19. The system of claim 14 , wherein the plurality of processing cores are further configured to shade the combined material parameters to produce an anti-aliased image.

20. The system of claim 14 , wherein the one or more samples are assigned to the plurality clusters based on depth values of the one or more samples.

Assignees

• Nvidia Corp

Inventors

• Crassin Cyril Jean-Francois
• Mcguire Morgan
• Lefohn Aaron Eliot
• Luebke David Patrick

Classifications

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06V10/763

  Non-hierarchical techniques, e.g. based on statistics of modelling distributions · CPC title

• G06F18/24137

  Distances to cluster centroïds · CPC title

• G06F18/23

  Clustering techniques · CPC title

• G06K9/6272

  Physics · mapped topic