On-site calibration of line-structured light vision sensor in complex light environments

Area of Science:

• Computer Vision
• Metrology
• Optical Sensing

Background:

• Line-structured light vision sensors are crucial for 3D reconstruction and measurement.
• Accurate calibration is essential for reliable sensor performance.
• Existing methods often require complex targets or procedures.

Purpose of the Study:

• To propose a novel, simplified calibration method for line-structured light vision sensors.
• To achieve high calibration accuracy comparable to traditional planar target methods.
• To reduce calibration complexity by using a movable parallel cylinder target.

Main Methods:

• Developed equations relating ellipse projections (from light stripe and cylinder intersection) to the light plane.
• Utilized perspective projection transformation principles.
• Applied constraint conditions (minor axis ellipse = cylinder diameter) to solve for the light plane equation.

Main Results:

• Achieved a calibration accuracy of 0.07 mm in physical experiments.
• Demonstrated the method's effectiveness within a 500 mm x 400 mm field of view and 700 mm measurement distance.
• Validated accuracy comparable to planar target calibration methods.

Conclusions:

• The proposed cylinder target calibration method is effective and accurate for line-structured light vision sensors.
• This method offers a simplified alternative to existing calibration techniques.
• The achieved accuracy supports its use in demanding metrology applications.