Mirror-based flexible one-step calibration method of light-sectioning system for surface topography

A flexible, easy-to-implement, one-step calibration method of the light-sectioning system is proposed for surface topography using a mirror adapter. The calibration method employs an arbitrary non-planar object with feature points as a target. With the assistance of the mirror adapter, the real and the reflected images of calibration are spatially separated and captured by a camera. It is developed as a direct calculation (DC) method and a virtual camera (VC) method to optimize the full parameters of the system. The DC method estimates the relative rotation and translation, respectively, and the VC method regards the two images as captured by two cameras in a physical sense through image flipping. The parameters of the laser plane can be solved accordingly. A numerical simulation is performed to analyze the precision of these two methods. A three-dimensional (3D) toy and a mechanical element are measured to verify the validity in translation experiments.