EPI-guided adaptive refocusing for high-SNR 3D measurement in a large-DOF MEMS-based structured light field

Area of Science:

• Optics and Photonics
• Metrology
• Computer Vision

Background:

• Structured light field (SLF) 3D measurement combines micro-electro-mechanical system (MEMS) projection with unfocused plenoptic cameras (UPC) for large depth of field (DOF) metrology.
• High laser coherence and small UPC sub-apertures cause speckle and low signal-to-noise ratio (SNR), degrading reconstruction accuracy.

Purpose of the Study:

• To propose and validate an EPI-guided adaptive refocusing model (EGARM) for enhancing 3D measurement accuracy in SLF systems.
• To address speckle noise and low SNR issues inherent in current SLF metrology techniques.
• To achieve stable and accurate 3D reconstruction over an extended depth range and field of view.

Main Methods:

• Employs phase-encoded epipolar plane image (EPI) analysis to extract phase features and estimate a depth-related slope factor.
• Utilizes the estimated slope factor to guide digital refocusing, aggregating multiple sub-aperture views from the 4D light field for speckle suppression.
• Introduces a refocused ray calibration and global geometric refinement framework to impose global geometric constraints.

Main Results:

• Demonstrates stable 3D reconstruction over an extended depth range of approximately 400-1856 mm.
• Achieves a significant expansion of the effective field of view (FOV) from 120 mm × 200 mm to 580 mm × 780 mm.
• Maintains a reconstruction accuracy consistently better than 0.45 mm across the measurement volume.

Conclusions:

• The proposed EGARM effectively suppresses speckle and improves SNR, leading to enhanced 3D reconstruction accuracy in SLF systems.
• The method enables stable and accurate 3D measurements over a large depth range and expanded FOV.
• EGARM represents a significant advancement in large-DOF metrology, overcoming limitations of previous techniques.