Simultaneous calibration method for rotary axis geometric errors in five-axis dispensing equipment based on monocular vision

Abstract Geometric error calibration of the rotary axis is an important and effective method for improving the five-axis dispensing equipment’s motion and processing accuracy. In most previous studies, simultaneously identifying the geometric errors of multiple rotation axes using traditional precision measuring instruments is often challenging. This paper proposes a simultaneous calibration model based on the AprilTag target visual framework. The model can synchronously identify the geometric errors of the rotation axis in the case of multi-axis linkage. The proposed kinematic calibration model avoids the singularity problem, and the calibration scheme is more economical and capable of unmanned and online calibration. Moreover, the proposed calibration model is compared against the traditional POE error model. In addition, simulation experiments are carried out using different regularization methods regarding the ill-posed problem in the recognition process under visual measurement conditions. Finally, the accuracy and feasibility of the proposed method are verified via experimental results on a five-axis dispensing equipment.