Shadow denoising in ray-tracing applications
In various examples, the actual spatial properties of a virtual environment are used to produce, for a pixel, an anisotropic filter kernel for a filter having dimensions and weights that accurately reflect the spatial characteristics of the virtual environment. Geometry of the virtual environment may be computed based at least in part on a projection of a light source onto a surface through an occluder, in order to determine a footprint that reflects a contribution of the light source to lighting conditions of the pixel associated with a point on the surface. The footprint may define a size, orientation, and/or shape of the anisotropic filter kernel and corresponding filter weights. The anisotropic filter kernel may be applied to the pixel to produce a graphically-rendered image of the virtual environment.
1. A method comprising:
determining a point corresponding to a surface in a virtual environment based at least on an interaction of a ray with the point in the virtual environment;
computing an intersection between an occluder of a light source and a three dimensional (3D) shape projecting from the point towards the light source in the virtual environment;
computing for a filter and based at least on the intersection, a filter geometry that corresponds to a cross-section of a projection of the intersection along a view vector; and
rendering an image corresponding to the virtual environment based at least on applying the filter to lighting condition data corresponding to the point.
2. The method of claim 1 , the intersection is between a portion of a virtual light path that extends from the point and a surface of the occluder that is on an opposite side of the occluder with respect to the light source.
3. The method of claim 1 , wherein the computing of the intersection is based at least on a distance between the point and the occluder.
4. The method of claim 1 , wherein the computing of the intersection is based at least on a normal vector of the surface at the point.
5. The method of claim 1 , wherein the computing of the intersection is based at least on a geometry of the light source, the geometry of the light source defining a shape of the intersection between the 3D shape and the occluder.
6. The method of claim 1 , wherein the filter geometry defines a first width of the filter along a first filter direction, and the applying of the filter uses the first width along the first filter direction and a second width along a second filter direction, the second width having a different magnitude than the first width.
7. The method of claim 1 , the computing of the intersection is based at least on a type of the light source.
8. A system comprising:
one or more processing units to execute instructions stored by one or more memory devices to cause the one or more processing units to perform operations including:
determining geometry of a three dimensional (3D) shape projecting from a point towards a light source in a virtual environment;
determining for a filter and using the geometry of the 3D shape, a filter geometry that corresponds to a cross-section of a projection of an intersection between an occluder and the 3D shape along a view vector; and
applying the filter to render data corresponding to the point to render an image corresponding to the virtual environment.
9. The system of claim 8 , wherein the filter geometry corresponds to a second intersection of the projection with a surface corresponding to the point.
10. The system of claim 8 , wherein the projection comprises a multi-dimensional shape.
11. The system of claim 8 , wherein a direction of the projection corresponds to the geometry of the 3D shape.
12. The system of claim 8 , wherein the determining of the filter geometry is based at least on a geometry of the light source, and the geometry of the light source defines a shape of the intersection between the 3D shape and the occluder.
13. The system of claim 8 , wherein the filter geometry defines a first width of the filter along a first filter direction, and the applying of the filter uses the first width along the first filter direction and a second width along a second filter direction, the second width having a different magnitude than the first width.
14. The system of claim 8 , the determining of the geometry is based at least on a type of the light source.
15. The system of claim 8 , wherein the light source is a rectangular light source and the 3D shape is a pyramid.
16. The system of claim 8 , wherein the light source is a directional light source and the 3D shape is a cone.
17. A processor comprising:
one or more circuits to render an image using a filter having a filter geometry computed based at least on a cross-section of a projection of an intersection between an occluder and a three dimensional (3D) shape along a view vector, the 3D shape projecting from a point towards a light source in a virtual environment, the projection being computed using geometry of the 3D shape.
18. The processor of claim 17 , wherein the intersection is between a portion of a virtual light path that extends from the point and a surface of the occluder that is on an opposite side of the occluder with respect to the light source.
19. The processor of claim 17 , wherein the filter geometry is computed based at least on a distance between the point and the occluder.
20. The processor of claim 17 , wherein the filter geometry is computed based at least on a normal vector of a surface at the point.