Phase Unwrapping Error Correction Based on Multiple Linear Regression Analysis.

Area of Science:

• 3D Measurement and Metrology
• Optical Engineering
• Computer Vision

Background:

• Fringe Projection Profilometry (FPP) is susceptible to Phase Unwrapping Errors (PUE) caused by noise and environmental factors.
• Existing PUE correction methods often fail to leverage the full spatial correlation within phase maps, limiting their effectiveness.

Purpose of the Study:

• To develop a novel and robust method for detecting and correcting Phase Unwrapping Errors (PUE) in Fringe Projection Profilometry (FPP).
• To improve the accuracy and reliability of 3D surface reconstruction from FPP data, especially in challenging regions.

Main Methods:

• Utilizing the low-rank property of unwrapped phase maps for PUE detection.
• Employing multiple linear regression to establish a regression plane and identify thick PUE regions based on tolerance.
• Applying an improved median filter to detect and mark random PUE positions.
• Implementing a final correction step for all identified PUE.

Main Results:

• The proposed method demonstrates high effectiveness and robustness in detecting and correcting PUE.
• Experimental results validate the accuracy and reliability of the new PUE correction technique.
• The method shows progressive capabilities in handling regions with abrupt or discontinuous phase changes.

Conclusions:

• The developed method offers a significant advancement in PUE correction for FPP.
• It provides a more comprehensive approach by utilizing global phase map information.
• This technique enhances the quality of 3D measurements obtained through FPP, particularly in complex scenarios.