Determining both surface position and orientation in structured-light-based sensing.

Area of Science:

• Computer Vision
• 3D Reconstruction
• Optical Metrology

Background:

• Accurate 3D shape description requires determining both position and orientation of surface elements.
• Traditional methods often rely on pattern element centroids, which are susceptible to distortions.

Purpose of the Study:

• To develop a structured light system for independent determination of position and orientation of surface elements.
• To utilize novel feature points (grid points) for enhanced accuracy and robustness.

Main Methods:

• Illumination with a pseudorandom pattern and selection of grid points between pattern elements as features.
• Exploitation of grid point symmetry for precise localization at subpixel accuracy.
• Color coding of grid points to aid in unique identification and orientation determination.

Main Results:

• Grid points are invariant to perspective distortion and allow independent extraction.
• Surface normals and 3D orientation are determined using the illuminated pattern's grid lines.
• Both position and orientation profiles are obtained from a single pattern illumination and image capture.

Conclusions:

• The proposed method accurately determines 3D position and orientation profiles using a novel feature point strategy.
• This approach offers improved robustness against distortions compared to traditional centroid-based methods.
• The technique enables efficient and accurate 3D shape analysis with minimal data acquisition.