Vision-based articulated machine pose estimation for excavation monitoring and guidance

Chen Feng, S. Dong, K. M. Lundeen, Y. Xiao, V. R. Kamat

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The pose of an articulated machine includes the position and orientation of not only the machine base (e.g., tracks or wheels), but also each of its major articulated components (e.g., stick and bucket). The ability to automatically estimate this pose is a crucial component of technical innovations aimed at improving both safety and productivity in many construction tasks. A computer vision based solution using a network of cameras and markers is proposed in this research to enable such a capability for articulated machines. Firstly, a planar marker is magnetically mounted on the end-effector of interest. Another marker is fixed on the jobsite whose 3D pose is pre-surveyed in a project coordinate frame. Then a cluster of at least two cameras respectively observing and tracking the two markers simultaneously forms a camera-marker network and transfers the end-effector's pose into the desired project frame, based on a pre-calibration of the relative poses between each pair of cameras. Through extensive sets of uncertainty analyses and field experiments, this approach is shown to be able to achieve centimeter level depth tracking accuracy within up to 15 meters with only two ordinary cameras (1.1 megapixel each) and a few markers, providing a flexible and cost-efficient alternative to other commercial products that use infrastructure dependent sensors like GPS. A working prototype has been tested on several active construction sites with positive feedback from excavator operators confirming the solution's effectiveness.

Original languageEnglish (US)
Title of host publication32nd International Symposium on Automation and Robotics in Construction and Mining
Subtitle of host publicationConnected to the Future, Proceedings
PublisherInternational Association for Automation and Robotics in Construction I.A.A.R.C)
ISBN (Electronic)9789517585972
StatePublished - Jan 1 2015
Event32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, ISARC 2015 - Oulu, Finland
Duration: Jun 15 2015Jun 18 2015

Other

Other32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, ISARC 2015
CountryFinland
CityOulu
Period6/15/156/18/15

Fingerprint

Excavation
Cameras
Monitoring
End effectors
Excavators
Computer vision
Global positioning system
Wheels
Innovation
Productivity
Calibration
Feedback
Sensors
Costs
Experiments

Keywords

  • Bundle adjustment
  • Camera-marker network
  • Excavation guidance
  • Pose estimation
  • Uncertainty analysis

ASJC Scopus subject areas

  • Building and Construction
  • Artificial Intelligence
  • Civil and Structural Engineering
  • Hardware and Architecture

Cite this

Feng, C., Dong, S., Lundeen, K. M., Xiao, Y., & Kamat, V. R. (2015). Vision-based articulated machine pose estimation for excavation monitoring and guidance. In 32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, Proceedings International Association for Automation and Robotics in Construction I.A.A.R.C).

Vision-based articulated machine pose estimation for excavation monitoring and guidance. / Feng, Chen; Dong, S.; Lundeen, K. M.; Xiao, Y.; Kamat, V. R.

32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, Proceedings. International Association for Automation and Robotics in Construction I.A.A.R.C), 2015.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Feng, C, Dong, S, Lundeen, KM, Xiao, Y & Kamat, VR 2015, Vision-based articulated machine pose estimation for excavation monitoring and guidance. in 32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, Proceedings. International Association for Automation and Robotics in Construction I.A.A.R.C), 32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, ISARC 2015, Oulu, Finland, 6/15/15.
Feng C, Dong S, Lundeen KM, Xiao Y, Kamat VR. Vision-based articulated machine pose estimation for excavation monitoring and guidance. In 32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, Proceedings. International Association for Automation and Robotics in Construction I.A.A.R.C). 2015
Feng, Chen ; Dong, S. ; Lundeen, K. M. ; Xiao, Y. ; Kamat, V. R. / Vision-based articulated machine pose estimation for excavation monitoring and guidance. 32nd International Symposium on Automation and Robotics in Construction and Mining: Connected to the Future, Proceedings. International Association for Automation and Robotics in Construction I.A.A.R.C), 2015.
@inproceedings{d9c902ee1e624bd5a4070f89097b4e8a,
title = "Vision-based articulated machine pose estimation for excavation monitoring and guidance",
abstract = "The pose of an articulated machine includes the position and orientation of not only the machine base (e.g., tracks or wheels), but also each of its major articulated components (e.g., stick and bucket). The ability to automatically estimate this pose is a crucial component of technical innovations aimed at improving both safety and productivity in many construction tasks. A computer vision based solution using a network of cameras and markers is proposed in this research to enable such a capability for articulated machines. Firstly, a planar marker is magnetically mounted on the end-effector of interest. Another marker is fixed on the jobsite whose 3D pose is pre-surveyed in a project coordinate frame. Then a cluster of at least two cameras respectively observing and tracking the two markers simultaneously forms a camera-marker network and transfers the end-effector's pose into the desired project frame, based on a pre-calibration of the relative poses between each pair of cameras. Through extensive sets of uncertainty analyses and field experiments, this approach is shown to be able to achieve centimeter level depth tracking accuracy within up to 15 meters with only two ordinary cameras (1.1 megapixel each) and a few markers, providing a flexible and cost-efficient alternative to other commercial products that use infrastructure dependent sensors like GPS. A working prototype has been tested on several active construction sites with positive feedback from excavator operators confirming the solution's effectiveness.",
keywords = "Bundle adjustment, Camera-marker network, Excavation guidance, Pose estimation, Uncertainty analysis",
author = "Chen Feng and S. Dong and Lundeen, {K. M.} and Y. Xiao and Kamat, {V. R.}",
year = "2015",
month = "1",
day = "1",
language = "English (US)",
booktitle = "32nd International Symposium on Automation and Robotics in Construction and Mining",
publisher = "International Association for Automation and Robotics in Construction I.A.A.R.C)",

}

TY - GEN

T1 - Vision-based articulated machine pose estimation for excavation monitoring and guidance

AU - Feng, Chen

AU - Dong, S.

AU - Lundeen, K. M.

AU - Xiao, Y.

AU - Kamat, V. R.

PY - 2015/1/1

Y1 - 2015/1/1

N2 - The pose of an articulated machine includes the position and orientation of not only the machine base (e.g., tracks or wheels), but also each of its major articulated components (e.g., stick and bucket). The ability to automatically estimate this pose is a crucial component of technical innovations aimed at improving both safety and productivity in many construction tasks. A computer vision based solution using a network of cameras and markers is proposed in this research to enable such a capability for articulated machines. Firstly, a planar marker is magnetically mounted on the end-effector of interest. Another marker is fixed on the jobsite whose 3D pose is pre-surveyed in a project coordinate frame. Then a cluster of at least two cameras respectively observing and tracking the two markers simultaneously forms a camera-marker network and transfers the end-effector's pose into the desired project frame, based on a pre-calibration of the relative poses between each pair of cameras. Through extensive sets of uncertainty analyses and field experiments, this approach is shown to be able to achieve centimeter level depth tracking accuracy within up to 15 meters with only two ordinary cameras (1.1 megapixel each) and a few markers, providing a flexible and cost-efficient alternative to other commercial products that use infrastructure dependent sensors like GPS. A working prototype has been tested on several active construction sites with positive feedback from excavator operators confirming the solution's effectiveness.

AB - The pose of an articulated machine includes the position and orientation of not only the machine base (e.g., tracks or wheels), but also each of its major articulated components (e.g., stick and bucket). The ability to automatically estimate this pose is a crucial component of technical innovations aimed at improving both safety and productivity in many construction tasks. A computer vision based solution using a network of cameras and markers is proposed in this research to enable such a capability for articulated machines. Firstly, a planar marker is magnetically mounted on the end-effector of interest. Another marker is fixed on the jobsite whose 3D pose is pre-surveyed in a project coordinate frame. Then a cluster of at least two cameras respectively observing and tracking the two markers simultaneously forms a camera-marker network and transfers the end-effector's pose into the desired project frame, based on a pre-calibration of the relative poses between each pair of cameras. Through extensive sets of uncertainty analyses and field experiments, this approach is shown to be able to achieve centimeter level depth tracking accuracy within up to 15 meters with only two ordinary cameras (1.1 megapixel each) and a few markers, providing a flexible and cost-efficient alternative to other commercial products that use infrastructure dependent sensors like GPS. A working prototype has been tested on several active construction sites with positive feedback from excavator operators confirming the solution's effectiveness.

KW - Bundle adjustment

KW - Camera-marker network

KW - Excavation guidance

KW - Pose estimation

KW - Uncertainty analysis

UR - http://www.scopus.com/inward/record.url?scp=84962840366&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84962840366&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84962840366

BT - 32nd International Symposium on Automation and Robotics in Construction and Mining

PB - International Association for Automation and Robotics in Construction I.A.A.R.C)

ER -