Optimizing Projector Pose Estimation for Selective Lighting in Optical See-Through Mixed Reality Systems

Projector-based selective lighting effectively addresses the limitations of optical see-through mixed reality systems caused by their inability to occlude real-world light. This study introduces a method for rapid and precise projector pose estimation to establish a hybrid system. The approach involves initial projector calibration using projected QR codes, followed by tracking and illuminating specific objects to achieve selective lighting. Experimental validation confirms the accuracy of the calibration process and demonstrates enhanced virtual object perception with the integration of an unilluminated aperture. Future research will focus on optimizing calibration and illumination techniques, as well as conducting user studies to assess improvements in virtual content perception across diverse scenarios.