Fourier domain ghost imaging with adaptive enhancement for scattering media

Area of Science:

• Optics and Photonics
• Remote Sensing
• Image Processing

Background:

• Atmospheric scattering degrades traditional optical imaging, limiting remote-sensing applications.
• Photon scattering, signal attenuation, and noise reduce image quality in challenging environments.

Purpose of the Study:

• To develop and validate a ghost imaging technique for improved image reconstruction in scattering media.
• To enhance image fidelity using adaptive post-processing algorithms.

Main Methods:

• Fourier-domain ghost imaging utilizing a digital micromirror device (DMD) for structured pattern generation.
• Frequency-domain correlation processing with a three-step phase-shifting algorithm for reconstruction.
• Multi-stage enhancement pipeline including adaptive denoising, deblurring, and contrast optimization.

Main Results:

• Recognizable image reconstruction achieved using only 25% of Fourier coefficients under high-scattering conditions.
• Adaptive enhancement algorithms improved peak signal-to-noise ratio (PSNR) by up to 30% compared to unprocessed reconstructions.
• Identified fundamental limitations in dense scattering regimes and computational trade-offs.

Conclusions:

• The proposed ghost imaging technique effectively reconstructs images in scattering environments.
• Adaptive post-processing significantly enhances image quality and signal-to-noise ratio.
• The study provides insights into ghost imaging system performance under challenging conditions.