Geometrical projection identity-based 3D reconstruction for fringe projection profilometry.

Area of Science:

• Computer Vision
• Geometric Computing
• Metrology

Background:

• 3D reconstruction is crucial for metrology and computer vision.
• Existing methods for 3D reconstruction have computational limitations.
• Focal Length Pinning (FPP) systems offer potential for improved reconstruction.

Purpose of the Study:

• To introduce novel geometric points and a geometric projection identity (GPI) for FPP systems.
• To develop new 3D reconstruction equations for common methods.
• To demonstrate the computational advantages of the GPI-based framework.

Main Methods:

• Identification of two special geometric points within an FPP system.
• Derivation of a geometric projection identity (GPI).
• Development of new 3D reconstruction equations (Ver3, Hor3, OptE3, HorVer4) based on GPI.
• Proof of equivalence between GPI-based and traditional reconstruction frameworks.

Main Results:

• The geometric projection identity (GPI) is revealed for FPP systems.
• New 3D reconstruction equations are provided for four established methods.
• The GPI-based framework is proven equivalent to traditional methods.
• Significantly faster point cloud computation is achieved using the GPI-based approach.

Conclusions:

• The geometric projection identity offers a more efficient approach to 3D reconstruction.
• The GPI-based method provides effective analysis of precision in metrology.
• This work advances 3D reconstruction techniques for FPP systems and metrology applications.