Related Experiment Video
Updated: Sep 13, 2025
09:41
Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping
Published on: April 21, 2023
1.7K
A Multi-View Three-Dimensional Scanning Method for a Dual-Arm Hand-Eye System with Global Calibration of Coded Marker
Tenglong Zheng1, Xiaoying Feng2, Siyuan Wang3
1School of Mechanical Engineering, Tiangong University, Tianjin 300387, China.
Micromachines
|July 30, 2025
Summary
This study introduces new global calibration methods for dual-arm hand-eye systems to improve 3D measurement accuracy. These techniques significantly reduce imaging errors in complex, noisy environments for intelligent manufacturing.
Area of Science:
- Robotics and Automation
- 3D Imaging and Metrology
- Computer Vision
Background:
- Accurate 3D measurement is crucial for intelligent manufacturing and collaborative robotics.
- Complex noise and robotic arm drift challenge existing hand-eye calibration methods.
- Traditional calibration techniques often lack robustness in dynamic, multi-view scenarios.
Purpose of the Study:
- To develop robust global calibration methods for dual-arm hand-eye systems in multi-view 3D imaging.
- To address stitching errors caused by robotic arm attitude drift and improve measurement accuracy.
- To provide a more precise and reliable solution for collaborative 3D measurement under challenging conditions.
Main Methods:
- A multi-view 3D scanning approach based on ICP (M3DHE-ICP) was developed, integrating multi-frequency heterodyne coding phase solution with ICP optimization.
- A global calibration method based on encoded marker points (GCM-DHE) was introduced, utilizing spatial geometry constraints and a dynamic tracking model.
- These methods were compared against traditional circular calibration plate methods.
Main Results:
- The M3DHE-ICP method reduced the average 3D imaging error to 0.082 mm (a reduction of 0.330 mm).
- The GCM-DHE method achieved an average imaging error of 0.100 mm, which is 0.456 mm lower than traditional methods.
- In engineering tests, scanning a vehicle's front mudguard yielded an average error of 0.085 mm with a standard deviation of 0.018 mm.
Conclusions:
- The proposed global calibration methods significantly enhance the accuracy and robustness of dual-arm hand-eye systems for multi-view 3D measurement.
- These advancements are vital for applications in intelligent manufacturing and multi-robot collaborative measurement systems.
- The developed techniques effectively mitigate errors from robotic arm drift and complex noise, offering superior performance over conventional approaches.
