Pose determination with semantic segmentation

Martin Hirzer (Inventor), Peter M. Roth (Inventor), Clemens Arth (Inventor), Vincent Lepetit (Inventor)

Research output: PatentResearch

Abstract

A method determines a pose of an image capture device. The method includes accessing an image of a scene captured by the image capture device. A semantic segmentation of the image is performed, to generate a segmented image. An initial pose of the image capture device is generated using a three-dimensional (3D) tracker. A plurality of 3D renderings of the scene are generated, each of the plurality of 3D renderings corresponding to one of a plurality of poses chosen based on the initial pose. A pose is selected from the plurality of poses, such that the 3D rendering corresponding to the selected pose aligns with the segmented image.
Original languageEnglish
Publication statusPublished - 2019

Fingerprint

Semantics

Cite this

@misc{cf567ccf859a4141a5d8160bb75197dc,
title = "Pose determination with semantic segmentation",
abstract = "A method determines a pose of an image capture device. The method includes accessing an image of a scene captured by the image capture device. A semantic segmentation of the image is performed, to generate a segmented image. An initial pose of the image capture device is generated using a three-dimensional (3D) tracker. A plurality of 3D renderings of the scene are generated, each of the plurality of 3D renderings corresponding to one of a plurality of poses chosen based on the initial pose. A pose is selected from the plurality of poses, such that the 3D rendering corresponding to the selected pose aligns with the segmented image.",
author = "Martin Hirzer and Roth, {Peter M.} and Clemens Arth and Vincent Lepetit",
year = "2019",
language = "English",
type = "Patent",

}

TY - PAT

T1 - Pose determination with semantic segmentation

AU - Hirzer, Martin

AU - Roth, Peter M.

AU - Arth, Clemens

AU - Lepetit, Vincent

PY - 2019

Y1 - 2019

N2 - A method determines a pose of an image capture device. The method includes accessing an image of a scene captured by the image capture device. A semantic segmentation of the image is performed, to generate a segmented image. An initial pose of the image capture device is generated using a three-dimensional (3D) tracker. A plurality of 3D renderings of the scene are generated, each of the plurality of 3D renderings corresponding to one of a plurality of poses chosen based on the initial pose. A pose is selected from the plurality of poses, such that the 3D rendering corresponding to the selected pose aligns with the segmented image.

AB - A method determines a pose of an image capture device. The method includes accessing an image of a scene captured by the image capture device. A semantic segmentation of the image is performed, to generate a segmented image. An initial pose of the image capture device is generated using a three-dimensional (3D) tracker. A plurality of 3D renderings of the scene are generated, each of the plurality of 3D renderings corresponding to one of a plurality of poses chosen based on the initial pose. A pose is selected from the plurality of poses, such that the 3D rendering corresponding to the selected pose aligns with the segmented image.

M3 - Patent

ER -