Adaptive compressed 3D ghost imaging based on the variation of surface normals

Area of Science:

• Computational imaging
• 3D reconstruction
• Optical metrology

Background:

• Computational ghost imaging with single-pixel detectors and photometric stereo enables 3D reconstruction.
• Large measurement counts and long imaging times hinder its practical application.
• The compressibility of surface normals is underexplored, impacting sampling efficiency.

Purpose of the Study:

• To develop an adaptive measurement method for 3D information acquisition.
• To improve the sampling efficiency of single-pixel imaging systems.
• To reduce the number of measurements and imaging time for 3D reconstruction.

Main Methods:

• Proposing an adaptive measurement strategy based on surface normal variations.
• Illuminating object surface regions with patterns of varying spatial resolutions.
• Utilizing computational ghost imaging and photometric stereo principles.

Main Results:

• The proposed method effectively reduces the number of required measurements.
• High-quality 3D images are preserved despite reduced sampling.
• Demonstrated feasibility through experimental validation.

Conclusions:

• Adaptive illumination based on surface normals enhances 3D imaging efficiency.
• The method overcomes limitations of traditional ghost imaging for 3D reconstruction.
• This approach offers a more practical solution for computational ghost imaging applications.