In the framework of the ESA-project “JERICOCalibration”, KRYPTON N.V. developed a compact and accurate optical measurement system [1]. The objective of this development was to obtain a lightweight system to correct for layout changes in robotic cells, due to thermal effects, collisions or rearrangement. The system consists of a single matrix camera and flat, passive targets. The camera is mounted on a robot end effector or tool, while the markers are attached to objects whose position and orientation in the robot’s environment have to be identified. The conducted research concentrated on the evaluation of different passive targets. A robust vision algorithm has been developed to find a unique solution for the target’s position and orientation, based on the ‘vanishing point method’. This algorithm and the targets were evaluated as far as calculation speed, noise sensitivity and accuracy is concerned. The innovative aspect in this development is the ability to determine an object in 6D (position and orientation) based on 1 single matrix camera and a small, flat, lowcost target. The calibrated camera has a measurement accuracy of 0.050 mm in a range of 100x100x100mm. When this camera is attached to the robot end effector, the measurement volume can be extended to the entire robot workspace. This concept was successfully tested with different matrix cameras and on different robotic installations. Both on a 6-axes industrial robot as well as on an 8-axes space robot, the system proved to be suitable for accurate environment calibration. Meanwhile, an industrial spin-off version has been elaborated, called GCS60, suitable for: • Localization of objects in the environment for offline programming or in changing environments (environment calibration) • Drift monitoring • Static performance testing (repeatability of the robot and backlash at joint level) • Encoder offset identification (robot calibration)
[1]
J F Quinet.
Calibration for offline programming purpose and its expectations
,
1995
.
[2]
Hendrik Van Brussel,et al.
Autonomous, calibrated robot arms take over the accurate execution of in orbit experiments
,
1996
.
[3]
Gideon P. Stein.
Internal Camera Calibration using Rotation and Geometric Shapes
,
1993
.
[4]
R.M. Haralick.
Monocular vision using inverse perspective projection geometry: analytic relations
,
1989,
Proceedings CVPR '89: IEEE Computer Society Conference on Computer Vision and Pattern Recognition.
[5]
A. Van den Bossche,et al.
Robot, Tool and Workcell Calibration: The Keys to Fully Integrate Off-Line Programming
,
1997
.