Search-free stereo-PMD: iterative reconstruction based on specular region consistency

Area of Science:

• Optics and Photonics
• Metrology
• Computer Vision

Background:

• Stereoscopic phase measuring deflectometry (stereo-PMD) is a key technique for specular surface measurement.
• Existing stereo-PMD methods face challenges with unique spatial point identification and the position of the slope sampling plane (SSP).
• The normal consistency criterion, while useful, does not always yield unique results.

Purpose of the Study:

• To develop an iterative reconstruction method for stereo-PMD.
• To eliminate the need for point cloud search and explicit positional consideration of the SSP.
• To improve the accuracy and robustness of specular surface reconstruction.

Main Methods:

• An iterative reconstruction approach for stereo-PMD was proposed.
• The normal consistency criterion was replaced by the constraint of specular regions consistency.
• The position of the SSP was dynamically optimized using this new constraint.

Main Results:

• The proposed method significantly improves low-order accuracy by over 80%.
• Noise robustness was enhanced by 33% compared to traditional methods.
• Eliminated the complex point cloud search and SSP positional considerations.

Conclusions:

• The novel iterative method using specular region consistency offers superior performance for stereo-PMD.
• This approach enhances measurement accuracy and robustness for specular surfaces.
• The method provides a more efficient and reliable reconstruction process.