Locally Adaptive Contrast Improvement for Underwater Image Enhancement

Abstract

The invention relates to a system and method for Locally Adaptive Contrast Improvement (LACI), designed to enhance the visual quality of underwater images. Underwater images often suffer from low contrast, color distortion, and noise due to light absorption, scattering, and the optical properties of water. This invention introduces a locally adaptive approach that divides the image into smaller regions, applying contrast and color enhancements independently to each area based on its specific characteristics. The system includes depth and distance compensation to adjust enhancements for objects at varying distances from the camera, restoring natural colors, particularly in the red and green spectrums. Additionally, it employs adaptive noise reduction techniques to minimize artifacts and noise amplification during the enhancement process. The LACI system is capable of real-time processing, making it suitable for applications such as live video feeds from autonomous underwater vehicles (AUVs), remotely operated vehicles (ROVs), and underwater cameras. By aligning image improvements with human visual perception, the system ensures that the final output appears natural and visually accurate. This invention is versatile and scalable, supporting a range of underwater imaging applications, including marine research, underwater exploration, industrial inspection, and recreational photography.

Specification

Description:The Locally Adaptive Contrast Improvement (LACI) system operates on the principle of localized image enhancement tailored specifically for underwater environments, where varying optical distortions affect different parts of an image. This principle ensures that each part of the image is enhanced based on its unique characteristics, rather than applying a one-size-fits-all global adjustment.
Here are the core principles behind the LACI system:
1. Local Contrast Adaptation
• Principle: Different regions of an underwater image suffer from varying degrees of contrast degradation due to factors like light absorption, scattering, and distance from the camera. The principle of local contrast adaptation involves dividing the image into smaller regions and applying different levels of contrast enhancement to each region independently.
• Goal: This localized approach ensures that dark or hazy areas receive more enhancement, while well-lit or clear areas remain unaffected, preventing overexposure or oversaturation.
2. Color Restoration
• Principle: Water absorbs different wavelengths of light at varying rates, with red light being absorbed the most, followed by green and blue. The principle of color restoration aims to compensate for this selective absorption by estimating the degree of color loss and restoring natural hues, particularly in the red and green spectrum.
• Goal: By applying this restoration locally, the system brings back true colors to objects and areas that would otherwise appear monochromatic or washed out, improving the color fidelity of the image.
3. Depth and Distance Compensation
• Principle: The deeper the object in the water or the farther away from the camera, the more light is absorbed, resulting in significant contrast loss and color distortion. The principle of depth and distance compensation involves analyzing each segment of the image to estimate its relative depth or distance from the camera and applying enhancement proportional to this distance.
• Goal: This allows the system to adjust the enhancement strength for deeper or more distant objects, ensuring a uniform and balanced result across the entire image.
4. Noise Reduction and Smoothing
• Principle: Enhancing contrast and brightness, especially in low-light conditions, can introduce noise and artifacts. The principle of adaptive noise reduction involves applying filters and smoothing techniques selectively to areas where noise may become an issue.
• Goal: The goal is to preserve the clarity and fine details of the image without introducing graininess or blurring textures, which can occur when noise is amplified during enhancement.
5. Perception-based Enhancement
• Principle: Human visual perception is sensitive to specific contrast levels and color variations. The system leverages perception-based algorithms that enhance the image based on how the human eye perceives changes in contrast and color. This principle ensures that the enhanced image looks natural and is not overly processed.
• Goal: The system's adjustments are made to optimize the visual quality of the image in a way that aligns with how humans perceive clarity, color, and contrast, making the images more realistic and visually appealing.
The system for Locally Adaptive Contrast Improvement (LACI) in Fig. 1 is designed to enhance underwater images by applying localized image processing techniques. It consists of several key components and processes that work together to improve contrast, restore color, reduce noise, and compensate for the optical challenges of underwater environments.
Key Components:
1. Image Segmentation Module:
o This module divides the underwater image into smaller local regions. Each region is analyzed separately to determine the specific enhancement needed based on the degree of light attenuation, scattering, and contrast reduction in that area.
2. Contrast Enhancement Module:
o The contrast enhancement module applies adaptive contrast adjustments to each segmented region. Instead of globally enhancing the entire image, this system improves contrast where needed, bringing out hidden details in darker or hazier areas without overexposing already clear regions.
3. Color Restoration Module:
o To address color loss in underwater images, especially in the red and green spectrum, the system includes a color restoration module. This module restores natural colors by adjusting color balance and compensating for color attenuation that occurs due to the water's selective absorption of light.
4. Noise Reduction and Smoothing Module:
o This module ensures that the enhancement process does not introduce unwanted artifacts or noise. By applying adaptive smoothing techniques, the system preserves the clarity of textures and details while minimizing the amplification of noise, especially in low-light or low-contrast areas.
5. Depth and Distance Compensation Module:
o Since light attenuation increases with depth and distance from the camera, this module estimates the depth and distance for different parts of the image. Based on these estimations, the system adjusts the strength of contrast and color enhancements to ensure uniform improvement across the image.
Workflow Process:
1. Input: The underwater image is first fed into the system.
2. Segmentation: The image is divided into localized regions for more precise processing.
3. Enhancement: Contrast and color restoration are applied adaptively based on the needs of each region.
4. Noise Reduction: Smoothing filters are applied to ensure the image remains clean and free of artifacts.
5. Output: A final enhanced image with improved contrast, restored colors, and reduced noise is produced, providing clearer and more detailed underwater visuals.
Deployment:
• The system can be deployed in real-time applications, such as for underwater photography, marine research, or on autonomous underwater vehicles (AUVs). It can also be used in post-processing for improving pre-captured underwater footage or images.
, C , C , Claims:
1. We claim that this method will improve the contrast of underwater images by dividing the image into localized regions.
2. We claim that the system compensates for light absorption and scattering in underwater images.
3. We claim that the invention reduces noise and visual artifacts in underwater images during contrast and color enhancement.
4. We claim that the system is scalable and reliable.

Documents