Robust Single-Shot 3D Reconstruction by Sparse-to-Dense Stereo Matching and Spline Function Based Parallax Modeling

Area of Science:

• Computer Vision
• 3D Imaging
• Optical Metrology

Background:

• High-accuracy, high-resolution 3D surface imaging is crucial for academic and industrial applications.
• Existing active stereo vision (ASV) methods face challenges in robustly reconstructing complex 3D shapes.

Purpose of the Study:

• To develop a novel sparse-to-dense structured light (SL) line-pattern based ASV approach.
• To achieve robust and high-resolution 3D shape reconstruction.

Main Methods:

• Proposed a sparse-to-dense stereo matching (SDSM) method for line clustering and matching.
• Designed a four-color SL line pattern with varying line densities.
• Developed a spline-function based parallax model (SFPM) for depth computation.

Main Results:

• The SDSM method effectively clusters and matches sparse and dense color lines.
• The SFPM accurately computes depths between matched color lines.
• Experimental results demonstrate superior robustness compared to existing methods, especially for complex shapes.

Conclusions:

• The proposed SDSM-SFPM ASV approach offers a robust solution for high-resolution 3D surface imaging.
• This technique significantly improves the reconstruction of complex 3D geometries.