Development of a Camera Localization System for Three-Dimensional Digital Image Correlation Camera Triangulation

Advancements in computer-vision methods have made stereophotogrammetry and three-dimensional digital image correlation (3D-DIC) attractive tools for Structural Health Monitoring (SHM). Nonetheless, optical approaches are not always well suited for quantitative measurement on large-scale structures due to the challenging calibration process of the cameras. In this study, a novel multi-sensor system is presented to compute the stereophotogrammetry system's extrinsic parameters. This system can extend 3D-DIC's capability and allow for easier calibration and measurement. The proposed system integrates a MEMS-based Inertial Measurement Unit (IMU) for determining the spatial orientation of the cameras (i.e., roll, pitch, and yaw angles) and a 77 GHz radar sensor for measuring the relative distance of the stereo cameras. The prototype's effectiveness is evaluated through laboratory and analytical tests. Its accuracy in calculating the extrinsic parameters of a stereophotogrammetry system is demonstrated by direct comparison with data obtained from a traditional calibration. Results show that the multi-sensor system can determine the separation angle between the two cameras with an error below 2% when the comparison is made.