Camera calibration method with focus-related intrinsic parameters based on the thin-lens model.

Area of Science:

• Computer Vision
• Metrology
• Optical Engineering

Background:

• High-magnification vision systems often suffer from inaccurate camera calibration and limited depth of field (DOF).
• Accurate measurement of small objects requires precise intrinsic parameter calibration and an extended DOF.

Purpose of the Study:

• To propose a novel camera calibration method addressing inaccurate calibration and small DOF issues in high-magnification vision systems.
• To achieve highly accurate 3D measurements of small objects with an extended DOF using a thin-lens model.

Main Methods:

• Development of a mathematical camera model incorporating focus-related intrinsic parameters.
• Initial camera calibration using the iterative radial alignment constraint (IRAC).
• Implementation of an optimization strategy for enhanced calibration accuracy.

Main Results:

• The proposed method significantly reduces root mean square errors in 3D coordinate measurements, from 22.02 μm to 1.66 μm as magnification increases.
• Demonstrated effectiveness, accuracy, and practicality through simulations and experimental validation.
• Achieved an extended DOF, crucial for measuring small objects with varying depths.

Conclusions:

• The thin-lens model-based calibration method enables accurate measurement of small objects.
• The technique effectively extends the depth of field (DOF) for vision systems.
• The proposed method offers a low calibration workload while enhancing measurement precision.