公开(公告)号：US10769849B2

公开(公告)日：2020-09-08

申请号：US15772740

申请日：2016-11-04

Applicant: INTEL CORPORATION

Inventor： Sridhar Uyyala ,  Bradley A. Jackson ,  Ignacio J. Alvarez ,  Deepak S. Vembar

IPC: G06T17/20 ,  G06T17/00 ,  G06T7/215 ,  G06T7/55 ,  G06T7/292 ,  G06T15/00 ,  G06T15/08

Abstract: Techniques for three-dimensional (3D) reconstruction of a dynamic scene as a set of voxels are provided. One technique includes: receiving, by a processor, image data from each of two or more spatially-separated sensors observing the scene from a corresponding two or more vantage points; fusing, by the processor, the image data into the set of voxels on a frame-by-frame basis; segmenting, by the processor, the image data into objects that constitute the scene; detecting, by the processor, which of the objects remain static from frame to frame, remaining ones of the objects being dynamic; filtering, by the processor, the set of voxels to remove those of the voxels corresponding to the static objects, to produce a dynamic subset of the voxels; and outputting, by the processor to a display device, those of the voxels corresponding to the dynamic objects (such as the dynamic subset) and not to the static objects.

公开(公告)号：US10580143B2

公开(公告)日：2020-03-03

申请号：US15772741

申请日：2016-11-04

Applicant: INTEL CORPORATION

Inventor： Sridhar Uyyala ,  Ignacio J. Alvarez ,  Bradley A. Jackson ,  Deepak S. Vembar

IPC: G06T7/215 ,  G06T17/10 ,  G06K9/00 ,  G06T1/20 ,  G06T13/40 ,  G06T17/20 ,  G06K9/62

Abstract: Techniques for high-fidelity three-dimensional (3D) reconstruction of a dynamic scene as a set of voxels are provided. One technique includes: receiving, by a processor, image data from each of two or more spatially-separated sensors observing the scene from a corresponding two or more vantage points; generating, by the processor, the set of voxels from the image data on a frame-by-frame basis; reconstructing, by the processor, surfaces from the set of voxels to generate low-fidelity mesh data; identifying, by the processor, performers in the scene from the image data; obtaining, by the processor, high-fidelity mesh data corresponding to the identified performers; and merging, by the processor, the low-fidelity mesh data with the high-fidelity mesh data to generate high-fidelity 3D output. The identifying of the performers includes: segmenting, by the processor, the image data into objects; and classifying, by the processor, those of the objects representing the performers.