Global Calibration of a Collaborative Multi-Line-Scan Camera Measurement System

Area of Science:

• Computer Vision
• Metrology

Background:

• Multi-line-scan cameras offer high-frequency sampling and wide fields of view for 3D measurement.
• Calibration challenges include scale ambiguity and parameter coupling due to 1D projection, limiting local optimization stability.

Purpose of the Study:

• To develop a global calibration method for multi-line-scan camera systems.
• To address scale ambiguity and parameter coupling for improved calibration consistency and stability.

Main Methods:

• A unified imaging and motion model incorporating physical scale constraints and structural priors.
• Introduction of geometric scale information into joint optimization.
• Parameter normalization and staged optimization for numerical stability.

Main Results:

• Stable convergence demonstrated under various perturbations (initialization, baseline deviation, noise).
• Achieved 3D reconstruction error below 0.67 mm.
• Demonstrated a convergence probability of at least 99.7%.

Conclusions:

• The proposed global calibration method effectively reduces uncertainty in multi-line-scan camera systems.
• The method supports high-precision online measurement and dynamic 3D perception.