US 11,748,949 B2
Object scanning using planar segmentation
Ke-Li Cheng, San Diego, CA (US); Kuang-Man Huang, Zhubei (TW); Michel Adib Sarkis, San Diego, CA (US); Gerhard Reitmayr, Del Mar, CA (US); and Ning Bi, San Diego, CA (US)
Assigned to QUALCOMM Incorporated, San Diego, CA (US)
Filed by QUALCOMM Incorporated, San Diego, CA (US)
Filed on May 13, 2022, as Appl. No. 17/744,484.
Application 17/744,484 is a continuation of application No. 16/998,975, filed on Aug. 20, 2020, granted, now 11,361,508.
Prior Publication US 2022/0343602 A1, Oct. 27, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 17/20 (2006.01); G06T 7/12 (2017.01); G06T 19/00 (2011.01)
CPC G06T 17/205 (2013.01) [G06T 7/12 (2017.01); G06T 19/006 (2013.01); G06T 2210/56 (2013.01)] 30 Claims
OG exemplary drawing
 
1. A method of generating one or more three-dimensional models, the method comprising:
obtaining a first frame of an object in a scene, wherein a portion of the object is positioned on a plane in the first frame;
detecting the plane in the first frame;
segmenting, based on the detected plane, the object from the plane in the first frame;
generating a three-dimensional (3D) model of the object based on segmenting the object from the plane;
generating an initial mesh for a portion of the 3D model of the object corresponding to the portion of the object positioned on the plane;
determining a boundary set including edges of the initial mesh that belong to at least one triangle of the initial mesh;
generating, using the boundary set, a refined mesh for the portion of the 3D model;
obtaining a second frame of the object in the scene, wherein the portion of the object is positioned on the plane in the second frame; and
determining whether to generate a 3D model of the object for the second frame based on performing plane detection or object tracking on the second frame.