公开(公告)号：US20220301175A1

公开(公告)日：2022-09-22

申请号：US17835538

申请日：2022-06-08

Applicant: Snap Inc.

Inventor： Piers Cowburn ,  David Li ,  Isac Andreas Müller Sandvik ,  Qi Pan ,  Andrew Tristan Spek

IPC: G06T7/11 ,  G06T7/12 ,  G06T7/194 ,  G06T7/143 ,  G06T7/174

Abstract: Methods, systems, and devices for generating a persistent world-space ground (or floor) segmentation map (or “texture”) for use in augmented or virtual reality 3D experiences.

公开(公告)号：US20220230351A1

公开(公告)日：2022-07-21

申请号：US17714720

申请日：2022-04-06

Applicant: Snap Inc.

Inventor： Patrick Fox-Roberts ,  Richard McCormack ,  Qi Pan ,  Edward James Rosten

IPC: G06T7/70 ,  G06T3/00 ,  G06T5/00

Abstract: The pose of a wide-angle image is determined by dewarping regions of the wide-angle image, determining estimated poses of the dewarped regions of the wide-angle image and deriving a pose of the wide-angle image from the estimated poses of the of the dewarped regions. The estimated poses of the dewarped regions may be determined by comparing features in the dewarped regions with features in prior dewarped regions from one or more prior wide-angle images, as well as by comparing features in the dewarped regions with features in a point cloud.

公开(公告)号：US11361493B2

公开(公告)日：2022-06-14

申请号：US17030755

申请日：2020-09-24

Applicant: Snap Inc.

Inventor： Piers Cowburn ,  David Li ,  Isac Andreas Müller Sandvik ,  Qi Pan

IPC: G06T7/11 ,  G06T15/04 ,  G06T17/20 ,  G06K9/62 ,  G06V30/262

Abstract: A semantic texture map system to generate a semantic texture map based on a 3D model that comprises a plurality of vertices that include coordinate that indicate positions of the plurality of vertices, a UV map, and a semantic segmentation image that comprises a set of semantic labels.

公开(公告)号：US20210375054A1

公开(公告)日：2021-12-02

申请号：US17083722

申请日：2020-10-29

Applicant: Snap Inc.

Inventor： Qi Pan ,  Edward James Rosten ,  Andrew Tristan Spek ,  Supannee Tanathong

IPC: G06T19/00 ,  G06T7/70

Abstract: An augmented reality device generates an updated position and orientation (pose) value by initially determining, using image-based processing, a pose estimate from a current image frame, a previous image frame and a previous pose. An IMU pose is then obtained from data generated by an inertial measurement unit including motion sensors, and transformation between the pose estimate and the IMU pose is determined. The transformation is applied to the pose estimate to generate a pose update. Media content is applied to a position within an image, the position being based on the pose update.