Fully Uncalibrated Image-Based Visual Servoing of 2DOFs Planar Manipulators With a Fixed Camera

Area of Science:

• Robotics
• Computer Vision
• Control Systems

Background:

• Traditional vision-based robotic control often requires camera calibration, which can be complex and prone to errors.
• Existing uncalibrated methods may offer limited convergence guarantees and impose constraints on camera and robot motion.
• Parameter estimation errors in camera calibration can significantly degrade robotic control performance.

Purpose of the Study:

• To develop fully uncalibrated visual servoing approaches for robotic manipulators.
• To eliminate the need for camera calibration and complex parameter estimation algorithms.
• To achieve robust and flexible robotic control with guaranteed convergence.

Main Methods:

• Proposed novel fully uncalibrated visual servoing strategies for position control of a 2-Degrees-of-Freedom (2-DOF) planar manipulator.
• Developed control laws that do not require camera parameter calibration or online parameter estimation.
• Introduced an approach that accommodates a general camera pose, relaxing the parallel plane constraint.

Main Results:

• Demonstrated exponential convergence of image position errors, irrespective of camera parameter uncertainties.
• Achieved superior convergence guarantees compared to existing uncalibrated methods that typically offer only asymptotic convergence.
• Validated the effectiveness and flexibility of the proposed methods through simulations and experimental results.

Conclusions:

• The presented fully uncalibrated visual servoing approaches offer a robust and simplified solution for robotic manipulator control.
• These methods enhance system design flexibility and improve control performance by removing reliance on calibration.
• The findings pave the way for more practical and adaptable vision-guided robotic systems.