SYSTEM AND METHOD OF PRESENTING VIEWS OF A VIRTUAL SPACE

Number: 20180114343
Published: 2018-04-26
Filed: 2016-10-21
Assignee: DISNEY ENTERPRISES, INC.
Inventors: Mitchell; Kenneth J.; Koniaris; Charalampos; Iglesias-Guitian; Jose Antonio; Moon; Bochang; Smolikowski; Eric
CPC: G06V10/764; G06F18/22; G06F18/24; G06T7/90; G06T11/10; G06T11/60; G06V10/955
Verdict: Low Notable software
Source: Google Patents · FreePatentsOnline

The keeper's note

Virtual-space view presentation.

Abstract

Views of a virtual space may be presented based on predicted colors of individual pixels of individual frame images that depict the views of the virtual space. Predictive models may be assigned to individual pixels that predict individual pixel colors of individual pixels at individual time points. Individual models may be updated and/or reprojected to other pixels based on comparisons of the predicted pixel colors and colors specified by in a raster input signal.

Background

FIELD OF THE DISCLOSURE

This disclosure relates to presenting views of a virtual space.BACKGROUND

Computer generated animations and/or video games may be generated from three-dimensional models of virtual spaces. The virtual space may be used to tell a story using one or more virtual objects that may be placed within the virtual space. The virtual objects may represent characters, topographical elements, and/or other objects within the virtual space. Virtual objects may be animated in accordance with a storyline, plot, action sequence, character dialogue, and/or other video content of the animated video. To generate an animation, two-dimensional images may be generated from the three-dimensional model at various points in time. The images may be compiled together to comprise frames of the video.

High quality real-time presentation of three-dimensional content generated with rasterization pipelines may exhibit temporal and spatial coherence. A classic problem in computer graphics is that rasterization pipelines may recompute each frame image from scratch, incurring a lot of repeated processing involving heavy geometry processing and/or shading stages. Although graphics cards continue to evolve their computational power, the increasing demand of more complex and sophisticated shading computations create a compromise between available computational resources and achievable spatial and temporal resolution. As a consequence, spatial and temporal resolution are of

Claims

1. A system configured for presenting views of a virtual space, the system comprising: one or more physical processors configured by machine-readable instructions to: obtain state information defining state of a virtual space, the virtual space including topography and virtual objects, such that the state of the virtual space at individual time points defines the topography, the virtual objects, and positions of virtual objects; determine time points associated within individual frame images of a video of the virtual space, individual frame images comprising individual pixels having pixel colors and pixel locations, individual frame images depicting individual views of the virtual space at individual ones of the associated time points, the time points including a first time point associated with a first frame image and a second time point associated with a second frame image; assign predictive models to individual pixels, individual predictive models being configured to predict individual pixel colors of individual pixels at individual time points, a first pixel being assigned a first model instance of a first predictive model and a second pixel being assigned a second model instance of a second predictive model; and for individual time points: predict, using assigned individual predictive models, individual pixel colors of the individual pixels, such that for the first time point the first model instance of the first predictive model predicts a first color for the first pixel in the first frame image and the second model instance of the second predictive model predicts a second color for the second pixel in the first frame image; obtain input pixel information, the input pixel information being determined from state information, the input pixel information specifying colors of individual pixels of individual frame images that depict a view of the virtual space at the individual time points, such that for the first time point the input pixel information specifies a third color for the first pixel in the first frame image and a fourth color for the second pixel in the first frame image; and for individual pixels in the individual frame images: determine differences between individual predicted pixel colors of the individual pixels and the individual specified pixel colors of the individual pixels included in the input pixel information, and differences between individual predicted pixel colors of neighboring pixels and the individual specified pixel colors of the individual pixels included in the input pixel information; responsive to the differences between individual predicted pixel colors of neighboring pixels and the individual specified pixel colors of the individual pixels included in the input pixel information being less than the differences between individual predicted pixel colors of the individual pixels and the individual specified pixel colors of the individual pixels included in the input pixel information, assign the individual pixel colors of the individual pixels in the individual frame images as the individual predicted pixel colors of the individual neighboring pixels, and reassign the individual pixels to other instances of individual predictive models assigned to the individual neighboring pixels, such that: for the first pixel at the first time point, based determining a first difference between the first color and the third color, and a second difference between the second color and the third color, and responsive to determining that the second difference is less than the first difference, the pixel color of the first pixel in the first frame image is assigned to the second color and the first pixel is reassigned to a third model instance of the second predictive model, and for the second pixel at the first time point, based determining a third difference between the second color and the fourth color, and a fourth difference between the first color and the fourth color, and responsive to determining that the third difference is less than the fourth difference, the pixel color of the second pixel in the first frame image is assigned to the second color and the assignment of the second model instance of the second predictive model to the second pixel is maintained; and update the individual predictive models assigned to the individual pixels based on the individual differences between individual assigned pixel colors and the individual specified pixel colors, such that: for the first pixel, the third model instance of the second predictive model is updated based on the second difference, and for the second pixel, the second model instance of the second predictive model is updated based on the third difference. 11. A method of presenting views of a virtual space, the method being implemented in a computer system comprising one or more physical processor and non-transitory electronic storage storing machine-readable instructions, the method comprising: obtaining state information defining state of a virtual space, the virtual space including topography and virtual objects, such that the state of the virtual space at individual time points defines the topography, the virtual objects, and positions of virtual objects; determining time points associated within individual frame images of a video of the virtual space, individual frame images comprising individual pixels having pixel colors and pixel locations, individual frame images depicting individual views of the virtual space at individual ones of the associated time points, the time points including a first time point associated with a first frame image and a second time point associated with a second frame image; assigning predictive models to individual pixels, individual predictive models being configured to predict individual pixel colors of individual pixels at individual time points, including a first pixel being assigned a first model instance of a first predictive model and a second pixel being assigned a second model instance of a second predictive model; and for individual time points: predicting, using assigned individual predictive models, individual pixel colors of the individual pixels, including for the first time point using the first model instance of the first predictive model to predict a first color for the first pixel in the first frame image and the second model instance of the second predictive model to predict a second color for the second pixel in the first frame image; obtaining input pixel information, the input pixel information being determined from state information, the input pixel information specifying colors of individual pixels of individual frame images that depict a view of the virtual space at the individual time points, including obtaining, for the first time point input pixel information that specifies a third color for the first pixel in the first frame image and a fourth color for the second pixel in the first frame image; and for individual pixels in the individual frame images: determining differences between individual predicted pixel colors of the individual pixels and the individual specified pixel colors of the individual pixels included in the input pixel information, and differences between individual predicted pixel colors of neighboring pixels and the individual specified pixel colors of the individual pixels included in the input pixel information; responsive to the differences between individual predicted pixel colors of neighboring pixels and the individual specified pixel colors of the individual pixels included in the input pixel information being less than the differences between individual predicted pixel colors of the individual pixels and the individual specified pixel colors of the individual pixels included in the input pixel information, assigning the individual pixel colors of the individual pixels in the individual frame images as the individual predicted pixel colors of the individual neighboring pixels, and reassigning the individual pixels to other instances of individual predictive models assigned to the individual neighboring pixels, such that: for the first pixel at the first time point, based determining a first difference between the first color and the third color, and a second difference between the second color and the third color, and responsive to determining that the second difference is less than the first difference, the pixel color of the first pixel in the first frame image is assigned to the second color and the first pixel is reassigned to a third model instance of the second predictive model, and for the second pixel at the first time point, based determining a third difference between the second color and the fourth color, and a fourth difference between the first color and the fourth color, and responsive to determining that the third difference is less than the fourth difference, the pixel color of the second pixel in the first frame image is assigned to the second color and the assignment of the second model instance of the second predictive model to the second pixel is maintained; and updating the individual predictive models assigned to the individual pixels based on the individual differences between individual assigned pixel colors and the individual specified pixel colors, such that: for the first pixel, updating the third model instance of the second predictive model based on the second difference, and for the second pixel, updating the second model instance of the second predictive model based on the third difference.