Clustered principal components for precomputed radiance transfer
First Claim
1. A graphics processing method for illuminating a geometrical object, the method comprising:
- compressing a quantity of per-vertex transfer matrices by, at least;
partitioning the per-vertex matrices into a quantity of clusters fewer in number than the quantity of per-vertex matrices, andapproximating each cluster as an affine subspace with a dimensionality dependent on at least the cluster;
computing a transfer matrix model for each cluster; and
producing an exit radiance for each cluster by applying the transfer matrix model to at least one input parameter the exit radiance producing shading for rendering and presentation on a display.
2 Assignments
0 Petitions
Accused Products
Abstract
Real-time processing includes per-point transfer matrices forming a high-dimensional surface signal that is compressed using clustered principal component analysis (CPCA). CPCA partitions multiple samples into fewer clusters, each cluster approximating the signal as an affine subspace. Further, source radiance is input to a processor, which approximates source radiance using spherical harmonic basis to produce a set of source radiance coefficients. A graphics processing unit (GPU) processes the source radiance coefficients through the transfer matrix model for each cluster. The result of such processing is the exit radiance, which parameterizes the radiance leaving the surface of the object at each point, thus producing the shading for each point of the virtual object in real time.
-
Citations
46 Claims
-
1. A graphics processing method for illuminating a geometrical object, the method comprising:
-
compressing a quantity of per-vertex transfer matrices by, at least; partitioning the per-vertex matrices into a quantity of clusters fewer in number than the quantity of per-vertex matrices, and approximating each cluster as an affine subspace with a dimensionality dependent on at least the cluster; computing a transfer matrix model for each cluster; and producing an exit radiance for each cluster by applying the transfer matrix model to at least one input parameter the exit radiance producing shading for rendering and presentation on a display. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 45, 46)
-
-
23. A computer-readable medium storing one or more instructions to be executed by one or more processors to render an illuminated virtual object, the one or more instructions causing the one or more processors to:
-
compress a quantity of per-point transfer matrices of the virtual object by, at least; partitioning the per-point transfer matrices into a quantity of clusters lesser in number than the quantity of per-point matrices, and approximating each cluster as an affine subspace with a dimensionality dependent on at least the cluster; determine a transfer matrix model for each cluster; and produce an exit radiance for each cluster by applying the transfer matrix model for each cluster to at least one input parameter. - View Dependent Claims (24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44)
-
Specification