Motion-induced error reduction by combining Fourier transform profilometry with phase-shifting profilometry.

Area of Science:

• Optical Metrology
• Computational Imaging
• 3D Reconstruction

Background:

• Motion-induced errors significantly challenge accurate 3D measurements in profilometry.
• Existing methods struggle with dynamic scenes and multiple objects.

Purpose of the Study:

• To develop a robust hybrid computational framework for reducing motion artifacts in 3D profilometry.
• To enable accurate measurement of multiple, isolated, rapidly moving objects.

Main Methods:

• A three-step hybrid approach combining single-shot Fourier Transform Profilometry (FTP) for relative phase maps and Phase-Shifting Profilometry (PSP) for absolute phase.
• Spatial phase unwrapping is employed in conjunction with FTP.
• Relative phase maps are aligned to an error-containing absolute phase map from PSP to generate final object-specific absolute phase maps.

Main Results:

• The proposed framework successfully extracts continuous relative phase maps for individual objects.
• An absolute phase map of the entire scene is obtained, despite inherent motion errors.
• Final absolute phase maps for each object are generated by referencing the scene's absolute phase map, effectively correcting for motion.
• Experimental validation confirms the framework's success in measuring multiple, isolated, rapidly moving objects.

Conclusions:

• The hybrid FTP-PSP computational framework effectively mitigates motion-induced errors in 3D profilometry.
• This method provides accurate 3D measurements for dynamic scenes involving multiple isolated objects.
• The approach offers a significant advancement for applications requiring high-speed 3D object measurement.