Hand-to-eye calibration of a robotic arm using point cloud registration method

The objective of this paper is to investigate a hand-eye calibration approach employing an eye-to-hand configuration, focusing on point cloud registration. In the initial segments of the paper, the paper delineate the core principles that underpin hand-eye calibration, as well as the various techniques that are predominantly utilized in this field. Furthermore, the article meticulously analyzes the limitations of the traditional methods and identifies potential areas where improvements can be made. Subsequently, it delves into the hand-eye calibration methodology grounded in 3D point cloud registration. This exploration includes an in-depth analysis of an enhanced version of the Super 4PCS Algorithm-based point cloud registration algorithm, which is a novel and effective method for point cloud registration. Additionally, the article discusses the intricacies of coordinate system transformation and provides a comprehensive explanation of how to solve the hand-eye calibration equation AX=XB using the navy method. Furthermore, the paper empirically validates the efficacy and precision of this method. This validation process involves a comparative analysis with other existing approaches. Lastly, it summarizes the research findings and discusses the strengths and limitations of the proposed method.