Underwater polarization correlation imaging based on backscatter-degraded speckle correction

Area of Science:

• Optical Imaging
• Image Processing
• Computational Imaging

Background:

• Underwater optical imaging suffers from image degradation due to backscatter noise.
• Existing methods struggle with effective backscatter suppression in turbid environments.

Purpose of the Study:

• To propose and validate an underwater polarization-correlated imaging method (ABS-UPGI) for backscatter noise correction.
• To enhance image quality and suppress backscatter in challenging underwater conditions.

Main Methods:

• Developed an ABS-UPGI method integrating polarization imaging with computational correlation imaging.
• Implemented backscatter degradation speckle correction to weaken speckle-noise correlation.
• Incorporated correlation-derived parameters into the polarization imaging model.

Main Results:

• ABS-UPGI demonstrated improved image reconstruction quality under medium-to-low sampling rates and strong scattering.
• Achieved a ~4 dB increase in peak signal-to-noise ratio (PSNR).
• Showcased a ~0.15 improvement in structural similarity index (SSIM) and 20%-25% overall image quality enhancement.

Conclusions:

• The ABS-UPGI method is effective and robust for suppressing backscatter in underwater environments.
• Offers a novel algorithmic approach for enhancing image quality in turbid waters.
• Validates the physical weakening of speckle-noise correlation for backscatter suppression.