Calibration of three-dimensional integrated sensors for improved system accuracy

When assembling multiple sensors on a three-dimensional integrated system, alignment problems usually exist along with the sensitive axes of the sensors, which subsequently cause system errors. This paper presents an optical calibration method to address this alignment problem. With the theorem of the least squares estimation, the misalignment errors can be identified by the three orthogonal optical alignments. The identified misalignment error is further compensated for the measurement of the system. A visible orthogonal reference system used for the optical alignments is also introduced, which helps realize a mounting alignment when installing the sensor system on a measured object. Experiments are performed with a MEMS-based azimuth-level detector to show effectiveness of the method.