Iterative camera calibration method using ray tracing and entrance pupil center constraints for phase measuring profilometry

To address the challenge of insufficient camera calibration accuracy in high-precision 3D measurement, this paper proposes an iterative method using ray tracing and entrance pupil center constraints. The entrance pupil center is estimated via ray tracing and introduced as a constraint, while outlier control points are removed using the 3<i>σ</i> criterion. Vision ray parameters are iteratively optimized with the Levenberg-Marquardt algorithm. Experimental results on a 120 mm metal bearing show that the proposed method significantly reduces reconstruction error: RMS decreases from 0.20 to 0.07 mm and PV from 1.39 to 0.76 mm, approaching the performance of coordinate measuring machines.