VDI/VDE 2634-2 and ISO 10360-13 performance evaluation tests, and systematic errors in structured light systems.

Structured light systems (SLSs) are commonly used for precision dimensional measurements. These are active systems in that the projected structured light pattern is used to establish correspondence between the projector and the camera(s) to provide more accurate 3D reconstruction as opposed to passive camera-based systems that rely on the user or features in images to identify corresponding points between the cameras. Errors in the calibration of the model parameters of an SLS can lead to systematic errors in dimensional measurements. In this paper, we explore the topic of performance testing of an SLS, specifically, the sensitivity of the length tests described in the VDI/VDE 2634-2 guideline and the ISO 10360-13 standard to the model parameters (such as intrinsic parameters of the camera and the projector, and camera-projector geometry parameters) of an SLS. The results are based on a hybrid approach of simulations performed using data obtained experimentally from an SLS mounted on a Cartesian coordinate measuring machine (CMM) which is used primarily to position the target at different points within the measurement volume. The results show that the length tests described in the ISO 10360-13 standard provide better sensitivity to the model parameters than the VDI/VDE 2634-2 guideline. However, even the ISO 10360-13 length tests do not detect all parameters with high sensitivity. A model-based approach in identifying length tests (i.e., the position and orientation of reference lengths) such as described in this paper provides better sensitivity to model parameters and therefore is more likely to detect systematic errors in SLSs. • Built a structured light system on a CMM and determined the length errors between points in a rectangular grid. • Assessed the sensitivity of lines defined in the VDI/VDE 2634-2 and ISO 10360-13 documents to systematic errors in SLSs. • Identified position/orientation of lines in the measurement volume that provides maximum sensitivity to systematic errors. • Demonstrated the superiority of our sensitivity analysis-based approach in detecting SLS systematic errors. [ABSTRACT FROM AUTHOR]