An Improvement of Pose Measurement Method Using Global Control Points Calibration

Area of Science:

• Computer Vision
• Metrology
• Robotics

Background:

• Vision-based pose measurement is crucial in complex environments.
• Existing methods suffer from significant errors that increase with measurement range.
• High accuracy is required for large-scale pose measurement applications.

Purpose of the Study:

• To propose a novel solution for reducing measurement errors in vision-based pose estimation.
• To enhance the accuracy of pose measurement across large ranges.
• Introduce Global Control Point Calibration (GCPC) as an optimized method.

Main Methods:

• Developed Global Control Point Calibration (GCPC) by optimizing existing visual pose measurement techniques.
• Differentiated measurement errors into control point error and control space error.
• Implemented global control points and object pose error calibration to process these errors.

Main Results:

• GCPC effectively eliminates control point errors and minimizes control space errors.
• Experimental results on moving targets show Root Mean Square (RMS) errors of 0.175° (yaw), 0.189° (pitch), and 0.159° (roll).
• Demonstrated stable and effective performance of the GCPC method.

Conclusions:

• GCPC significantly improves the accuracy of vision-based pose measurement, especially over large ranges.
• The proposed method offers a robust solution for applications demanding high precision pose estimation.
• GCPC provides a stable and effective approach to overcome limitations of traditional methods.