Determination of the object surface function by structured light: application to the study of spinal deformities

Area of Science:

• Optics and Photonics
• Computer Vision
• Biomedical Engineering

Background:

• Structured light projection is a key technique for 3D surface shape determination.
• Accurate correspondence matching between projected grids and object surfaces is crucial but challenging.
• Existing methods often require precise calibration of camera and projector positions.

Purpose of the Study:

• To develop a novel, efficient procedure for resolving grid-to-object point correspondences in structured light projection.
• To enable accurate 3D surface reconstruction without stringent calibration requirements.
• To apply the method for non-invasive medical diagnostics, specifically for spinal deformities.

Main Methods:

• Utilizes three images of a projected grid: two for calibration on flat surfaces (before and behind the object) and one on the object itself.
• Employs a calibration process independent of the precise relative positioning of the camera, projector, and object.
• Analyzes the projected grid pattern on the object to derive its surface function.

Main Results:

• Successfully establishes efficient correspondence matching between projected grid knots and object surface points.
• Achieves accurate 3D surface reconstruction by analyzing the projected grid on the object post-calibration.
• Demonstrates suitability for objects without discontinuities or significant depth gradients.

Conclusions:

• The proposed method offers an efficient and less calibration-dependent approach to structured light 3D surface reconstruction.
• It is particularly effective for smooth surfaces and has potential applications in non-invasive medical imaging.
• Quantitative diagnosis of spinal deformities like scoliosis can be achieved by analyzing back surface function symmetry differences.