Joint optimization of an ArUco-pillar composite target and a fixed reference for high-precision camera calibration...

High-precision camera calibration is essential for ensuring reliable environmental perception and precise visual servo control in robotic systems. Traditional target calibration methods have exhibited limited robustness and high sensitivity to environmental disturbances, primarily due to the dispersion of the coordinate system resulting from multiple target poses. An ArUco-pillar composite target is developed to improve the efficiency of feature point extraction and to enrich spatial information. At the same time, a fixed reference is integrated to reduce the dimensionality of the calibration parameter space. These innovations jointly contribute to enhanced calibration accuracy. Furthermore, a reward feedback-driven sampling strategy is introduced to improve the robustness of the calibration outcomes further. The proposed method is validated through both simulation and real-world experiments, in which the simulation data contain varying levels of uniformly distributed noise. Experimental results demonstrate that, compared with conventional methods, the proposed approach achieves significantly higher calibration accuracy and robustness.