Multiview phase-measuring profilometry calibration using a static target

Abstract

Phase-measuring profilometry (PMP), widely used for capturing micro-scale geometry, requires a phaseto- height calibration by illuminating multiple phase shifts on a static target of surfaces with different heights. However, it is only valid for a single-view system. When having more than one view, we must calibrate each camera and projector’s parameters separately. This additional calibration requires capturing many checkerboard images by physically rotating the target. It is cumbersome and not easily implementable within a tiny room inside a scanning instrument. In this thesis, we aim to combine these two modalities of the phase-to-height and the camera/projector parameter calibrations, enabling multiview PMP with high accuracy. To this end, we devise a novel compact, static calibration target with planar surfaces of different orientations with fiducial markers and a joint multiview optimization scheme of the projectors and the cameras. First, we automatically detect the markers to estimate plane equation parameters of different surface orientations. We then solve homography matrices of multiple planes using bundle adjustment. Given unwrapped phase measurement, we estimate the focal length, principal point, lens distortion, rotation, and translation parameters of every camera and projector. It allows us to measure depths through ray-plane intersections and combine them as a 3D object. Our method does not require physical interaction with the target during calibration. Only one static scene is required for calibration. Results validate that our calibration method enables us to combine multiview PMP measurements with high accuracy.