Refining Visually Detected Object poses

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Automated industrial assembly today require that the 3D position and orientation (hereafter ''pose`) of the objects to be assembled are known precisely. Today this precision is mostly established by a dedicated mechanical object alignment system. However, such systems are often dedicated to the particular object and in order to handle the demand for flexibility, there is an increasing demand for avoiding such dedicated mechanical alignment systems. Rather, it would be desirable to automatically locate and grasp randomly placed objects from tables, conveyor belts or even bins with a high accuracy that enables direct assembly. Conventional vision systems and laser triangulation systems can locate randomly placed known objects (with 3D CAD models available) with some accuracy, but not necessarily a good enough accuracy. In this paper, we present a novel method for refining the pose accuracy of an object that has been located based on the appearance as detected by a monocular camera. We illustrate the quality of our refinement method experimentally.
Original languageEnglish
Title of host publicationProceedings of the International Symposium on Robotics
Number of pages6
Publication date2010
Publication statusPublished - 2010
EventInternational Symposium on Robotics - Munich, Germany
Duration: 7. Jun 20109. Jun 2010
Conference number: 41

Conference

ConferenceInternational Symposium on Robotics
Number41
CountryGermany
CityMunich
Period07/06/201009/06/2010

Fingerprint

Refining
Bins
Triangulation
Computer aided design
Cameras
Lasers

Keywords

  • pose refinement

Cite this

Holm, P., & Petersen, H. G. (2010). Refining Visually Detected Object poses. In Proceedings of the International Symposium on Robotics
Holm, Preben ; Petersen, Henrik Gordon. / Refining Visually Detected Object poses. Proceedings of the International Symposium on Robotics. 2010.
@inproceedings{e1aefaf07de811df846a000ea68e967b,
title = "Refining Visually Detected Object poses",
abstract = "Automated industrial assembly today require that the 3D position and orientation (hereafter ''pose`) of the objects to be assembled are known precisely. Today this precision is mostly established by a dedicated mechanical object alignment system. However, such systems are often dedicated to the particular object and in order to handle the demand for flexibility, there is an increasing demand for avoiding such dedicated mechanical alignment systems. Rather, it would be desirable to automatically locate and grasp randomly placed objects from tables, conveyor belts or even bins with a high accuracy that enables direct assembly. Conventional vision systems and laser triangulation systems can locate randomly placed known objects (with 3D CAD models available) with some accuracy, but not necessarily a good enough accuracy. In this paper, we present a novel method for refining the pose accuracy of an object that has been located based on the appearance as detected by a monocular camera. We illustrate the quality of our refinement method experimentally.",
keywords = "pose refinement",
author = "Preben Holm and Petersen, {Henrik Gordon}",
year = "2010",
language = "English",
booktitle = "Proceedings of the International Symposium on Robotics",

}

Holm, P & Petersen, HG 2010, Refining Visually Detected Object poses. in Proceedings of the International Symposium on Robotics. International Symposium on Robotics, Munich, Germany, 07/06/2010.

Refining Visually Detected Object poses. / Holm, Preben; Petersen, Henrik Gordon.

Proceedings of the International Symposium on Robotics. 2010.

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

TY - GEN

T1 - Refining Visually Detected Object poses

AU - Holm, Preben

AU - Petersen, Henrik Gordon

PY - 2010

Y1 - 2010

N2 - Automated industrial assembly today require that the 3D position and orientation (hereafter ''pose`) of the objects to be assembled are known precisely. Today this precision is mostly established by a dedicated mechanical object alignment system. However, such systems are often dedicated to the particular object and in order to handle the demand for flexibility, there is an increasing demand for avoiding such dedicated mechanical alignment systems. Rather, it would be desirable to automatically locate and grasp randomly placed objects from tables, conveyor belts or even bins with a high accuracy that enables direct assembly. Conventional vision systems and laser triangulation systems can locate randomly placed known objects (with 3D CAD models available) with some accuracy, but not necessarily a good enough accuracy. In this paper, we present a novel method for refining the pose accuracy of an object that has been located based on the appearance as detected by a monocular camera. We illustrate the quality of our refinement method experimentally.

AB - Automated industrial assembly today require that the 3D position and orientation (hereafter ''pose`) of the objects to be assembled are known precisely. Today this precision is mostly established by a dedicated mechanical object alignment system. However, such systems are often dedicated to the particular object and in order to handle the demand for flexibility, there is an increasing demand for avoiding such dedicated mechanical alignment systems. Rather, it would be desirable to automatically locate and grasp randomly placed objects from tables, conveyor belts or even bins with a high accuracy that enables direct assembly. Conventional vision systems and laser triangulation systems can locate randomly placed known objects (with 3D CAD models available) with some accuracy, but not necessarily a good enough accuracy. In this paper, we present a novel method for refining the pose accuracy of an object that has been located based on the appearance as detected by a monocular camera. We illustrate the quality of our refinement method experimentally.

KW - pose refinement

M3 - Article in proceedings

BT - Proceedings of the International Symposium on Robotics

ER -

Holm P, Petersen HG. Refining Visually Detected Object poses. In Proceedings of the International Symposium on Robotics. 2010