INAM-Based Image-Adaptive 3D LUTs for Underwater Image Enhancement

Area of Science:

• Computer Vision
• Image Processing
• Underwater Imaging

Background:

• Underwater images suffer from poor visibility due to scattering and absorption.
• Existing image enhancement methods often struggle with preserving details and color fidelity.

Purpose of the Study:

• To enhance underwater images by improving the performance of adaptive three-dimensional lookup tables (3D LUTs).
• To address the limitations of Instance Normalization in feature extraction for underwater image enhancement.

Main Methods:

• Proposing an Instance Normalization Adaptive Modulator (INAM) to amplify pixel bias and improve feature distinction.
• Integrating the INAM into a learning framework for image-adaptive 3D LUTs.
• Utilizing bias amplification to make edge information more distinguishable in features.

Main Results:

• The proposed INAM significantly improves the performance of adaptive 3D LUTs for underwater image enhancement.
• Enhanced distinction of edge information leads to superior image quality.
• Experimental results demonstrate the effectiveness of the INAM-enhanced method over existing approaches.

Conclusions:

• The INAM is a novel and effective approach to enhance underwater image quality.
• Adaptive 3D LUTs integrated with INAM offer a substantial improvement in underwater image enhancement.
• This method provides a promising direction for future research in underwater computer vision.