- Number
- 11710272
- Published
- 2023-07-25
- Filed
- 2021-03-24
- Assignee
- Disney Enterprises, Inc.
- Inventors
- Coffey; Dane M. et al.
- CPC
- G06T15/08; G06T17/00; G06T17/20; G06T7/593; G06T17/10; G06T15/20; H04N5/2224
- Verdict
- Low Notable software
- Source
- Google Patents · FreePatentsOnline
The keeper's note
Volumetric capture/rendering technique from depth data (granted).
Abstract
An image processing system includes a computing platform having processing hardware, a display, and a system memory storing a software code. The processing hardware executes the software code to receive a digital object, surround the digital object with virtual cameras oriented toward the digital object, render, using each one of the virtual cameras, a depth map identifying a distance of that one of the virtual cameras from the digital object, and generate, using the depth map, a volumetric perspective of the digital object from a perspective of that one of the virtual cameras, resulting in multiple volumetric perspectives of the digital object. The processing hardware further executes the software code to merge the multiple volumetric perspectives of the digital object to form a volumetric representation of the digital object, and to convert the volumetric representation of the digital object to a renderable form.
Background
BACKGROUND (1) The types of three-dimensional digital objects used in content production pipelines can be incredibly complex. For example, digital models of characters and other artistic assets are typically in the form of meshes having many parts and including upwards of millions of polygons. The complexity of digital assets, while not free of challenges for pre-rendered content, are typically manageable due to the advantage of lengthy offline renders on large compute farms. However, real-time rendering of such high complexity digital assets on mobile devices or in game engines is impracticable. As a result, those high complexity digital assets must be significantly simplified for their use on mobile platforms or in game engines to be possible. (2) In the conventional art, simplification of a film quality digital asset often includes remodeling the asset by one or more artists to reduce its complexity. Due to its intense reliance on human participation, this manual remodeling is both undesirably costly and time consuming. Consequently, there is a need in the art for a substantially automated solution for converting high complexity film quality digital objects to digital objects that are simple enough to be rendered on mobile devices and in game engines. While there exist techniques to perform automatic decimation on models, most implementations act on source geometry that lack important render-time details (e.g. application of displacement maps, procedurals, etc.).
Claims
1. An image processing system comprising: a computing platform including a processing hardware, a display, and a system memory storing a software code; the processing hardware configured to execute the software code to: surround a digital object with a plurality of virtual cameras oriented toward the digital object, the digital object being a three-dimensional (3D) digital object, wherein the plurality of virtual cameras is a predetermined plurality, and wherein surrounding the digital object with the plurality of virtual cameras includes: generating at least one closed surface surrounding the digital object; casting a plurality of occlusion rays from each of a plurality of points on a surface of the digital object toward a surface of the at least one closed surface, resulting in a plurality of intersections; clustering the plurality of intersections, based on the plurality of virtual cameras and a surface density of the plurality of intersections on the at least one closed surface, to identify a plurality of camera locations each corresponding respectively to one of the plurality of virtual cameras; and generating the plurality of virtual cameras at the plurality of camera locations; render, using each of the plurality of virtual cameras, a depth map identifying a distance of each of the plurality of virtual cameras from the digital object; generate, using the depth map, a volumetric perspective of the digital object from a perspective of each of the plurality of virtual cameras, resulting in a corresponding plurality of volumetric perspectives of the digital object; merge the plurality of volumetric perspectives of the digital object to form a volumetric representation of the digital object; and convert the volumetric representation of the digital object to a renderable form. ||
10. A method for use by an image processing system including a computing platform having a processing hardware, a display, and a system memory storing a software code, the method comprising: surrounding, by the software code executed by the processing hardware, a digital object with a plurality of virtual cameras oriented toward the digital object, the digital object being a three-dimensional (3D) digital object, wherein the plurality of virtual cameras is a predetermined plurality, and wherein surrounding the digital object with the plurality of virtual cameras includes: generating at least one closed surface surrounding the digital object; casting a plurality of occlusion rays from each of a plurality of points on a surface of the digital object toward the at least one closed surface, resulting in a plurality of intersections; clustering the plurality of intersections, based on the plurality of virtual cameras and a surface density of the plurality of intersections on the at least one closed surface, to identify a plurality of camera locations each corresponding respectively to one of the plurality of virtual cameras; and generating the plurality of virtual cameras at the plurality of camera locations; rendering, by the software code executed by the processing hardware and using each of the plurality of virtual cameras, a depth map identifying a distance of each of the plurality of virtual cameras from the digital object; generating, by the software code executed by the processing hardware and using the depth map, a volumetric perspective of the digital object from a perspective of each of the plurality of virtual cameras, resulting in a plurality of volumetric perspectives of the digital object; merging, by the software code executed by the processing hardware, the corresponding plurality of volumetric perspectives of the digital object to form a volumetric representation of the digital object; and converting, by the software code executed by the processing hardware, the volumetric representation of the digital object to a renderable form. ||
17. The method claim 10, wherein the digital object is represented by a first mesh representation having a first mesh element count, and wherein the renderable form is represented by a second mesh representation having a second mesh element count less than the first mesh element count.