Large object vision measurement system based on a fixed connection between a camera and a total station telescope

Area of Science:

• Optics and Photonics
• Computer Vision
• Metrology

Background:

• Accurate three-dimensional (3D) measurement of large objects, especially their shape and color, is crucial in various industrial and scientific fields.
• Existing stereo vision methods often face challenges with scale, precision, and robustness when applied to large-scale metrology tasks.
• The need for adaptable and precise measurement systems for distant, large objects remains a significant challenge in computer vision and engineering.

Purpose of the Study:

• To develop a freely adjustable vision measurement system capable of precisely measuring the 3D shape and color of large objects at far ranges.
• To introduce a high-precision calibration method utilizing non-parametric camera models for enhanced accuracy.
• To implement a point cloud-based 3D reconstruction technique integrated with the proposed camera models.

Main Methods:

• Development of a vision measurement system by integrating a camera with a total station telescope using a fixed mount.
• Proposal of a non-parametric camera model-based calibration method for high-precision camera parameter estimation.
• Integration of non-parametric camera models with a point cloud-based method for 3D reconstruction of target objects.

Main Results:

• Experimental validation demonstrated the system's high precision and robustness in far-range measurements.
• Indoor testing at 8m showed a maximum spatial distance error of less than 0.24mm (relative error < 0.09%) on a calibration board.
• Outdoor testing at 20m yielded an average distance measurement error of 5.6mm (average relative error 0.12%) over an 11m x 4m region.

Conclusions:

• The developed freely adjustable vision measurement system effectively addresses the challenges of large-object metrology at far distances.
• The proposed non-parametric calibration and point cloud reconstruction methods ensure high accuracy and reliability.
• The system's performance in both indoor and outdoor scenarios validates its practical applicability in precise large-scale 3D measurements.