Phase Error Reduction for a Structured-Light 3D System Based on a Texture-Modulated Reprojection Method.

Area of Science:

• Optical Metrology
• 3D Imaging
• Computer Vision

Background:

• Fringe projection profilometry (FPP) is a precise 3D optical measurement technique.
• Texture and defocus at reflective edges cause significant phase errors in FPP.
• Existing models struggle with non-ideal pixel brightness at texture boundaries.

Purpose of the Study:

• To develop a novel strategy for accurate 3D reconstruction using FPP.
• To address and eliminate phase errors caused by scene texture and system defocus.
• To improve the precision of FPP measurements at challenging texture edges.

Main Methods:

• Theoretical analysis of the non-pixel-wise phase propagation model for texture edges.
• Proposal of a reprojection strategy based on scene texture modulation.
• Generation of a reprojection weight mask and modulated stripe patterns for error elimination.
• Fusion of coarse and refined phase maps for accurate phase generation.

Main Results:

• The proposed reprojection strategy effectively mitigates texture edge effects.
• Experimental validation on smooth, depth, and curved surfaces demonstrated significant improvements.
• Root mean square error (RMSE) at texture edges decreased by 53.32%.

Conclusions:

• The reprojection strategy significantly enhances the accuracy of FPP.
• The method is effective in eliminating depth errors at texture edges.
• This approach advances precision reconstruction in scenarios with complex surface textures.