High-speed absolute three-dimensional shape measurement using three binary dithered patterns.

Area of Science:

• Computer Vision
• Computational Geometry
• Optical Metrology

Background:

• Accurate 3D reconstruction is crucial for various applications.
• Existing high-speed 3D measurement techniques often require complex patterns or hardware.
• There is a need for efficient and robust methods for dynamic 3D shape analysis.

Purpose of the Study:

• To develop a high-speed 3D geometry reconstruction method.
• To utilize simple 1-bit binary dithered patterns for 3D shape measurement.
• To establish robust pixel correspondence for accurate 3D reconstruction.

Main Methods:

• Employing three encoded 1-bit binary dithered patterns for structured light projection.
• Implementing a novel pixel matching strategy in object space, not image space, for robust correspondence.
• Developing an algorithm for absolute 3D geometry reconstruction from the captured patterns.

Main Results:

• Successfully reconstructed high-speed absolute 3D geometry.
• Demonstrated the feasibility of achieving high-speed 3D shape measurement using 1-bit patterns.
• Validated the robustness of the proposed correspondence matching algorithm through experiments.

Conclusions:

• The proposed method enables efficient and high-speed 3D shape measurement.
• The use of 1-bit binary patterns simplifies the system while maintaining accuracy.
• The object-space matching approach enhances the robustness of 3D reconstruction.