Industrial camera model positioned on an effector for automated tool center point calibration

Area of Science:

• Robotics
• Computer Vision
• Industrial Automation

Background:

• Accurate calibration of robotic manipulators is crucial for industrial applications.
• Existing methods for tool center point (TCP) calibration have limitations in precision.

Purpose of the Study:

• To present a novel, full camera model for enhanced robotic manipulator TCP calibration.
• To simulate the capture of calibration patterns for automated TCP calibration processes.

Main Methods:

• Development of a full camera model integrated with an industrial robotic arm's end-effector.
• Creation of an experimental robot stand for generating reference datasets.
• Comparison of the proposed full camera model against the traditional camera obscura model.

Main Results:

• The full camera model achieved significantly greater accuracy than state-of-the-art methods.
• Absolute error was reduced by up to seven times compared to the camera obscura model.
• Achieved an error of 3.65 pixels for small rotations (up to 5°) and translations (up to 20 mm).

Conclusions:

• The proposed full camera model offers a substantial improvement in TCP calibration accuracy for robotic manipulators.
• The solution is applicable to real-life industrial processes, enhancing automation reliability.
• This innovative approach provides a significant benefit over existing calibration techniques.