Ghost imaging in dynamic underwater environments by dual-measurement intensity correction

Area of Science:

• Optics and Photonics
• Underwater Imaging
• Computational Imaging

Background:

• Water-column scattering significantly degrades underwater imaging quality, particularly in dynamic conditions.
• Computational ghost imaging offers partial mitigation through inherent averaging, leading to Gaussian-like noise distributions.
• Existing deep-learning methods often rely on synthetic noise, which doesn't reflect real-world underwater imaging challenges.

Purpose of the Study:

• To develop and validate an intensity-correction protocol for enhancing underwater computational ghost imaging.
• To address the limitations of clean ground-truth images in real-world underwater scenarios.
• To improve the fidelity of reconstructed images obtained through computational ghost imaging in scattering media.

Main Methods:

• Implemented an intensity-correction protocol involving averaging two independent measurements.
• Developed an on-the-fly noise variance estimation technique.
• Applied a diagonal Wiener filter to correct detected intensities before image reconstruction.

Main Results:

• Simulations demonstrated the effectiveness of the proposed intensity-correction protocol.
• Experimental results confirmed a marked improvement in underwater computational ghost imaging reconstruction quality.
• The method successfully corrected intensity distortions caused by water-column scattering.

Conclusions:

• The presented intensity-correction protocol significantly enhances underwater computational ghost imaging performance.
• This approach provides a practical solution for improving image quality in challenging underwater environments.
• The findings are crucial for advancing underwater optical imaging technologies.