Face augmentation in video
US-12165275-B2 · Dec 10, 2024 · US
US9761039B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-9761039-B2 |
| Application number | US-201514665099-A |
| Country | US |
| Kind code | B2 |
| Filing date | Mar 23, 2015 |
| Priority date | Jul 22, 2014 |
| Publication date | Sep 12, 2017 |
| Grant date | Sep 12, 2017 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
A hybrid rendering method includes subdividing a two-dimensional (2D) frame into virtual tiles (hereinafter, referred to as ‘tiles’); classifying the tiles into first tiles which include an object having reflective or transparent properties and second tiles which do not include the object having reflective or transparent properties; storing geometric information of a primitive in the first tiles in a graphics processing unit (GPU) internal memory; and generating a first image by rendering one or more of the first tiles via ray tracing using the geometric information, generating a second image by rendering one or more of the second tiles via rasterization, and outputting a final image by merging the first and second images.
Opening claim text (preview).
What is claimed is: 1. A hybrid rendering method comprising: subdividing a two-dimensional (2D) frame into tiles; classifying the tiles into first tiles comprising an object having reflective or transparent properties and second tiles lacking the object having the reflective or transparent properties; storing geometric information of a primitive in the first. tiles in a graphics processing unit (GPU) internal memory; and generating a first image by rendering one or more of the first tiles via ray tracing using the geometric information of the primitive of the first tiles, generating a second image by rendering one or more of the second tiles via rasterization, and outputting a final image by merging the first and second images, wherein the classifying of the tiles comprises: extracting properties of primitives in the tiles; and determining, among the tiles, a tile including a primitive having one of a reflectivity and a refractive index greater than 0 as the first tile, by using the extracted properties of the primitives, wherein the extracting of the properties of the primitives in the tiles comprises: performing binning to determine a tile in which each of the primitives is included from among the tiles; generating information about mapping between the respective tiles and the primitives; and acquiring list of the primitives in each of the tiles by using the mapping information, to extract properties of the primitives in the list of primitives. 2. The hybrid rendering method of claim 1 , wherein the geometric information of the primitive of the first tiles comprises any one or any combination of depth information, a normal vector, and position information of the primitive of the first tiles. 3. The hybrid rendering method of claim 1 , further comprising: storing the geometric information of the primitive of the first tiles; fetching the geometric information of the primitive of the first tiles from a graphics processing unit (GPU) external memory; and copying the fetched geometric information into the GPU internal memory. 4. The hybrid rendering method of claim 3 , wherein the geometric information of the primitive of the first tiles comprises any one or any combination of depth information, a normal vector, and position information of the primitive, and wherein the GPU external memory is a G-buffer. 5. The hybrid rendering method of claim 1 , further comprising: creating a tile list comprising a classification result; and converting a primitive of the first tiles into a fragment by reading the tile list, wherein the classification result is used in the storing of the geometric information of the primitive. 6. A hybrid rendering apparatus comprising: a tile subdivision unit configured to subdivide a two-dimensional (2D) frame into tiles; a tile classification unit configured to classify the tiles into first tiles comprising an object having reflective or transparent properties and second tiles lacking the object having reflective or transparent properties; and a rendering unit configured to generate a first image by rendering one or more of the first tiles via ray tracing using geometric information of a primitive of the first tiles, to generate a second image by rendering one or more of the second tiles via rasterization, and to output a final image by merging the first and second images, wherein the tile classification unit extracts properties of primitives in the tiles and determines, among the tiles, a tile including a primitive having one of a reflectivity and a refractive index greater than 0 as the first tile by using the extracted properties of the primitives, wherein the tile classification unit comprises: a mapping unit configured to perform binning to determine a tile in which each of the primitives is included from among the tiles and to generate information about mapping between the respective tiles and the primitives; and a determination unit configured to acquire a list of the primitives in each of the tiles by using the mapping information, extract properties of the primitives in the list of primitives, and to determine, among the tiles, a tile including one primitive having one of a reflectivity and a refractive index greater than 0 as the first tile, by using the properties of the primitives. 7. The hybrid rendering apparatus of claim 6 , wherein the geometric information of the primitive of the first tiles comprises any one or any combination of depth information, a normal vector, and position information of the primitive of the first tiles. 8. The hybrid rendering apparatus of claim 6 , wherein the rendering via ray tracing is performed together with shading in a device for performing the shading. 9. The hybrid rendering apparatus of claim 6 , further comprising: a fetching unit configured to store the geometric information of the primitive in the first tiles in a graphics processing unit (GPU) internal memory and to fetch the geometric information of the primitive in the first tiles from a GPU external memory and copy the fetched geometric information into the GPU internal memory. 10. The hybrid rendering apparatus of claim 9 , wherein the geometric information of the primitive of the first tiles comprises any one or any combination of depth information, a normal vector, and position information of the primitive, and wherein the GPU external memory is a G-buffer. 11. The hybrid rendering apparatus of claim 9 , further comprising a rasterizer, wherein the tile classification unit creates a tile list comprising a classification result, wherein the rasterizer converts a primitive of the first tiles into a fragment by reading the tile list, and wherein the fetching unit uses the classification result contained in the tile list. 12. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 1 . 13. A hybrid rendering method comprising: subdividing a two-dimensional (2D) frame into tiles; generating mapping information of the tiles and primitives included in the tiles; classifying the tiles into first tiles and second tiles based on a characteristic of the primitives included in the tiles; storing geometric information of primitives included in the first tiles in a graphics processing unit (GPU) internal memory; performing rendering of the first tiles via ray tracing based on the geometric information of the primitives included in the first tiles thereby generating a first image and performing rendering via rasterization of the second tiles thereby generating a second image; and outputting a final image by merging the first and the second images, wherein the classifying of the tiles comprises: extracting properties of primitives in the tiles; and determining, among the tiles, a tile including a primitive having one of reflectivity and a refractive index greater than 0 as the first tile, by using the extracted properties of the primitives, wherein the extracting of the properties of the primitives in the tiles comprises: performing, binning, to determine a tile in which each of the primitives is included from among the tiles; generating information about mapping between the respective tiles and the primitives; and acquiring a list of the primitives in each, of the tiles by using the mapping information, to extract properties of the primitives in the list of primitives. 14. The hybrid rendering method of claim 13 , wherein the tiles are classified into the first tiles by determining whether the first tiles include at least one primitive having one of a reflectivity and
Filling planar surfaces by adding surface attributes, e.g. adding colours or textures · CPC title
Ray-tracing · CPC title
General purpose rendering architectures · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.