Hybrid-quality-guided phase fusion model for high dynamic range 3D surface measurement by structured light technology

Area of Science:

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

Background:

• Measuring 3D shapes of high dynamic range (HDR) surfaces using structured light is challenging.
• Current methods fuse multi-exposure images but use simplistic phase selection criteria, leading to inaccuracies due to noise and non-linear camera responses.

Purpose of the Study:

• To develop a novel hybrid-quality-guided phase fusion (HPF) model for accurate 3D shape measurement of HDR surfaces.
• To overcome limitations of existing phase selection methods in structured light 3D reconstruction.

Main Methods:

• A hybrid-quality measure is proposed to comprehensively evaluate phase quality.
• Initial phases are weighted and fused using the hybrid-quality measure for improved accuracy.
• The HPF model integrates multi-group image fusion with enhanced quality assessment.

Main Results:

• The HPF model provides a more accurate phase selection compared to traditional methods.
• Experimental validation demonstrates the model's effectiveness in reconstructing complete and accurate 3D shapes of HDR surfaces.
• Improved robustness against image noise, camera non-linearities, and local reflections.

Conclusions:

• The proposed HPF model significantly enhances the accuracy and completeness of 3D surface reconstruction for HDR objects.
• This method offers a robust solution for challenging 3D metrology applications involving high dynamic range surfaces.