Quantification of error from system and environmental sources in Optical See-Through head mounted display calibration methods

摘要

A common problem with Optical See-Through (OST) Augmented Reality (AR) is misalignment or registration error with the amount of acceptable error being heavily dependent upon the type of application. Approximation methods, driven by user feedback, have been developed to estimate the necessary corrections for alignment errors. These calibration methods, however, are susceptable to induced error from system and environmental sources, such as human alignment error. The proposed research plan is intended to further the development of accurate and robust calibration methods for OST AR systems by quantifying the impact of specific factors shown to contribute to calibration error. An important aspect of this research will be to develop methods for examining each factor in isolation in order to determine the independent error contribution of each source. This will facilitate the establishment of acceptable thresholds for each type of error and be a meaningful step toward defining quality metrics for OST AR calibration techniques.