Joint Static Calibration for Multiview Phase-Measuring Profilometry
摘要
Conventional phase-measuring profilometry (PMP) calibration methods are primarily designed for single-view systems. Extending PMP to multiview configurations typically requires separate geometric calibrations for each camera and projector, often relying on repeated manual repositioning of a calibration target. This decoupled, two-stage procedure is labor-intensive and prone to error accumulation and local minima. We propose a unified framework for static calibration of multiview PMP systems that jointly optimizes geometric and phase-to-height parameters in a single formulation. Our approach is built on two key components: (a) a compact multi-planar calibration target with embedded fiducial markers, which removes the need for target motion, and (b) a joint optimization strategy that simultaneously estimates intrinsic, extrinsic, and lens distortion parameters of all cameras and projectors from a single static scene. The proposed pipeline automatically detects planar features, constructs multiview homographies via a target-aware bundle adjustment, and incorporates unwrapped phase measurements to refine system parameters. This fully automated approach significantly reduces user intervention while improving stability across views. Experimental results demonstrate that our method provides accurate calibration and supports consistent fusion of multiview phase measurements.