Device-agnostic calibration for multi-modal MR systems

The unique advantages of mixed reality (MR) in facilitating the interaction between virtual and physical spaces have garnered widespread favor in the industrial and medical fields. However, the current spatial calibration methods for MR glasses often suffer from hardware-specific limitations, and their accuracy fails to meet the requirements of practical applications. To address this issue, we propose a rotational calibration framework that enables universal coordinate mapping between any MR glasses and external tracking systems. The core innovation of this framework lies in its hardware-independent rotational calibration protocol: by performing constrained rotational movements in multiple spatial configurations and integrating optical patterns with infrared reflective markers, it achieves sub-millimeter precision alignment between any MR glasses and motion capture systems. We conducted experiments using the HoloLens and OptiTrack, and quantitative evaluations across various calibration scenarios confirmed the method’s robustness and broad applicability beyond specific hardware combinations.