Intrinsic parameter-free calibration of FPP using a ray phase mapping model

Area of Science:

• Optical Engineering
• 3D Metrology
• Computer Vision

Background:

• Fringe Projection Profilometry (FPP) is a key technique for 3D surface reconstruction.
• Traditional FPP methods require complex calibration of intrinsic camera parameters, including lens distortion.
• These calibration steps can be time-consuming and system-specific.

Purpose of the Study:

• To introduce a novel Ray Phase Mapping (RPM) model for FPP.
• To eliminate the need for intrinsic parameter calibration in FPP systems.
• To develop a universal and accurate method for 3D mapping across different FPP configurations.

Main Methods:

• The proposed Ray Phase Mapping (RPM) model characterizes the imaging system using independent rays for each pixel.
• It associates these rays with the projected phase in the illumination field for efficient 3D mapping.
• Two distinct loss functions are employed to optimize camera ray parameters and mapping coefficients.

Main Results:

• The RPM model successfully avoids complex, imaging-specific modeling of lens layout and distortion.
• Experiments demonstrate high accuracy in 3D mapping for wide-angle lens FPP, telecentric lens FPP, and MEMS-based FPP systems.
• The method proves feasible and effective across diverse FPP hardware.

Conclusions:

• The Ray Phase Mapping (RPM) model offers a simplified and accurate approach to 3D reconstruction in FPP.
• Its ability to bypass intrinsic parameter calibration makes it a versatile solution for various FPP systems.
• This universal approach has significant potential for advancing 3D metrology applications.