Simultaneous Robot⁻World and Hand⁻Eye Calibration without a Calibration Object.

Area of Science:

• Robotics
• Computer Vision
• Calibration Techniques

Background:

• Accurate robot-world and hand-eye calibration are crucial for robotic applications, especially in constrained environments like mobile or medical robotics.
• Traditional calibration methods often require precise, sterile, or expensive objects and sufficient workspace, limiting their applicability.
• The need for robust calibration methods that are adaptable to various robotic systems and environments is growing.

Purpose of the Study:

• To propose an extended robot-world and hand-eye calibration method for evaluating camera-robot transformation relationships.
• To develop a method applicable in scenarios where specialized calibration objects or ample movement space are unavailable.
• To enhance the accuracy and effectiveness of robotic calibration for diverse applications.

Main Methods:

• Established a mathematical model using the Kronecker product for robot-gripper-to-camera and robot-base-to-world rigid transformations.
• Introduced sparse bundle adjustment for optimizing robot-world and hand-eye calibration parameters.
• Validated the approach through experiments using real datasets and a Denso robot.

Main Results:

• Achieved a translation relative error of less than 8/10,000 for rigid transformations within a 1.3 m × 1.3 m × 1.2 m movement range.
• Demonstrated a distance measurement mean error of 0.13 mm after three-dimensional reconstruction.
• Validated the effectiveness and accuracy of the proposed extended calibration method.

Conclusions:

• The proposed extended calibration method provides a highly accurate and effective solution for robot-world and hand-eye transformations.
• This approach is suitable for mobile and medical robotics, overcoming limitations of traditional methods regarding equipment and space.
• The method offers precise 3D reconstruction capabilities with minimal error, enhancing robotic system performance.