Detail publikace

Camera calibration method of optical system for large field measurement of hot forgings in heavy industry

Originální název

Camera calibration method of optical system for large field measurement of hot forgings in heavy industry

Anglický název

Camera calibration method of optical system for large field measurement of hot forgings in heavy industry

Jazyk

en

Originální abstrakt

During the manufacturing process of heavy forgings, simple contact measuring techniques are still used to check the dimensions, therefore an optical measuring system is in demand. In this paper, a camera calibration method for the passive measuring system, which is being developed in collaboration with an industrial partner, is proposed. Our approach is based on space resection and works with robust coded targets, which are distributed in the field of view. The coordinates of targets are measured using TRITOP (GOM) measuring system. This solution allows to build a large calibration field, without a need of large calibration objects. The camera calibration works in 2 steps - at first, the intrinsic parameters of the camera, including lens distortion, are calibrated. These parameters are considered as stable, due to the use of special camera covers. Multi-image version of the calibration method and dense field of calibration targets are used. The second step is performed from every image and employs a single-image extrinsic camera parameters calibration method. Only a few coded calibration targets, mounted on stable objects in the scene, are required. The calibration method was tested in industrial conditions. The method showed great results, the average reprojection error was under 0.1 px. The effect of thermally affected zone on the calibration process is discussed.

Anglický abstrakt

During the manufacturing process of heavy forgings, simple contact measuring techniques are still used to check the dimensions, therefore an optical measuring system is in demand. In this paper, a camera calibration method for the passive measuring system, which is being developed in collaboration with an industrial partner, is proposed. Our approach is based on space resection and works with robust coded targets, which are distributed in the field of view. The coordinates of targets are measured using TRITOP (GOM) measuring system. This solution allows to build a large calibration field, without a need of large calibration objects. The camera calibration works in 2 steps - at first, the intrinsic parameters of the camera, including lens distortion, are calibrated. These parameters are considered as stable, due to the use of special camera covers. Multi-image version of the calibration method and dense field of calibration targets are used. The second step is performed from every image and employs a single-image extrinsic camera parameters calibration method. Only a few coded calibration targets, mounted on stable objects in the scene, are required. The calibration method was tested in industrial conditions. The method showed great results, the average reprojection error was under 0.1 px. The effect of thermally affected zone on the calibration process is discussed.

BibTex


@inproceedings{BUT158410,
  author="Jakub {Hurník} and Aneta {Zatočilová} and David {Paloušek}",
  title="Camera calibration method of optical system for large field measurement of hot forgings in heavy industry",
  annote="During the manufacturing process of heavy forgings, simple contact measuring techniques are still used to check the dimensions, therefore an optical measuring system is in demand. In this paper, a camera calibration method for the passive measuring system, which is being developed in collaboration with an industrial partner, is proposed. Our approach is based on space resection and works with robust coded targets, which are distributed in the field of view. The coordinates of targets are measured using TRITOP (GOM) measuring system. This solution allows to build a large calibration field, without a need of large calibration objects. The camera calibration works in 2 steps - at first, the intrinsic parameters of the camera, including lens distortion, are calibrated. These parameters are considered as stable, due to the use of special camera covers. Multi-image version of the calibration method and dense field of calibration targets are used. The second step is performed from every image and employs a single-image extrinsic camera parameters calibration method. Only a few coded calibration targets, mounted on stable objects in the scene, are required. The calibration method was tested in industrial conditions. The method showed great results, the average reprojection error was under 0.1 px. The effect of thermally affected zone on the calibration process is discussed.",
  address="SPIE",
  booktitle="Optical Measurement Systems for Industrial Inspection XI",
  chapter="158410",
  doi="10.1117/12.2527693",
  edition="Proceedings of SPIE",
  howpublished="online",
  institution="SPIE",
  number="-",
  year="2019",
  month="june",
  pages="1--12",
  publisher="SPIE",
  type="conference paper"
}