Range-intensity-profile prior dehazing method for underwater range-gated imaging.

Area of Science:

• Optics and Photonics
• Image Processing
• Underwater Imaging

Background:

• Backscatter significantly degrades image quality in underwater environments.
• Range-gated imaging systems are susceptible to backscatter noise.
• Existing dehazing methods often struggle with complex underwater conditions.

Purpose of the Study:

• To develop a robust method for mitigating backscatter in single gated images.
• To estimate scene depth and remove water backscatter at varying depths.
• To enhance image quality for underwater range-gated imaging applications.

Main Methods:

• A range-intensity-profile prior dehazing method is proposed.
• Scene depth is estimated using the prior of target intensity distribution with range intensity profiles.
• Depth transmission and depth-noise maps are calculated from estimated scene depth.
• Image restoration is achieved by subtracting the depth-noise map and dividing by depth transmission.

Main Results:

• The proposed method effectively removes water backscatter at different depths.
• Restored images show significant improvement in quality.
• Peak signal-to-noise ratio (PSNR) of dehazing images increased up to a doubled value.
• The method performs well even with minor inaccuracies in depth estimation.

Conclusions:

• The proposed range-intensity-profile prior method offers effective backscatter mitigation in underwater range-gated imaging.
• This approach enhances image quality and signal-to-noise ratio, crucial for underwater exploration and analysis.
• The method demonstrates robustness and provides a significant improvement over existing techniques.