Hand-eye calibration using convex optimization

Hand-eye calibration is a very important task in robotics. Many algorithms have been proposed for it, but they almost apply the L₂ optimization, which is usually in the form of nonlinear optimization. In this paper, we propose new hand-eye calibration algorithms using convex optimization, and it can be solved in the form of a global linear optimization without starting values. Experiments with both simulated and real data are performed to test our algorithms. The experimental results show the robustness and validity of our algorithms. Considering both the computing errors and the time consuming, our algorithm based on quaternions is a good option for real applications.