Single-shot 3D shape measurement based on RGB dot patterns and stereovision

Area of Science:

• Computer Vision
• 3D Metrology
• Optical Measurement

Background:

• Traditional stereo matching algorithms in one-shot structured light 3D measurement suffer from low accuracy and efficiency.
• Limitations in projector and imaging resolution hinder precise 3D shape reconstruction.
• RGB dots projection offers potential for improved stereo matching in single-shot 3D measurement.

Purpose of the Study:

• To develop an accurate and efficient single-shot 3D measurement method using RGB dots projection.
• To address the limitations of traditional stereo matching algorithms in 3D reconstruction.
• To enhance the accuracy and reliability of 3D surface reconstruction.

Main Methods:

• Proposed a single-shot 3D measurement technique utilizing RGB dots projection.
• Implemented line clustering stereo matching for coarse matching of segmented red points.
• Employed slope differenced distribution and area constrained erosion for robust RGB dot extraction.
• Utilized a model correction method to restore and constrain un-imaged patterns.

Main Results:

• Achieved a high accuracy of approximately 0.089mm in 3D reconstruction.
• Demonstrated superior performance compared to traditional disparity and RGB line methods.
• Successfully reconstructed 3D surfaces with enhanced precision.

Conclusions:

• The proposed line clustering stereo matching and model correction method significantly improves single-shot 3D measurement accuracy.
• This approach offers a viable solution for high-precision 3D surface reconstruction.
• The method shows promise for applications requiring accurate and efficient 3D measurements.