Simultaneous Calibration of Odometry and Head-Eye Parameters for Mobile Robots with a Pan-Tilt Camera

Area of Science:

• Robotics
• Computer Vision
• Sensor Fusion

Background:

• Accurate robot navigation relies on precise odometric and head-eye parameters.
• Existing calibration methods are often separate, increasing time and inconvenience.
• Current simultaneous methods do not support mobile robots with pan-tilt cameras.

Purpose of the Study:

• To develop a comprehensive simultaneous calibration method for odometric and head-eye parameters.
• To address the limitations of previous methods for mobile robots with pan-tilt camera systems.
• To improve the accuracy and efficiency of robot calibration.

Main Methods:

• Synchronized acquisition of visual features (from chessboard/natural scenes) and odometry measurements.
• Iterative adjustment of odometric and head-eye parameters until convergence.
• Application of nonlinear optimization for enhanced accuracy.

Main Results:

• Successfully calibrated both odometric and head-eye parameters simultaneously for a pan-tilt camera system.
• Demonstrated a method applicable to common mobile robot configurations.
• Overcame the limitation of not estimating the z-axis translation parameter in previous approaches.

Conclusions:

• The proposed method offers a full simultaneous calibration for mobile robots with pan-tilt cameras.
• This approach significantly reduces calibration time and enhances operational precision.
• Enables more accurate robot navigation in diverse environments.