Underwater Image Color Balancing with Water Tint Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing color balancing techniques in photography fail to accurately correct images taken underwater, often removing the blue tint from water, resulting in images that do not appear as if they were taken underwater.
Innovation Solution
An image editing application that adjusts colors in a multi-step process, protecting colors near the water's color while removing undesirable tints from other objects, using a YIQ color space to adjust pixels away from the water's color with reduced adjustments near the water's color, and applying inverse gamma adjustments to maintain original luminance levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If previous color balancing techniques are applied to remove the blue tint from water, then the water color is corrected to appear neutral, but the image no longer appears to be taken underwater
Solution Approach 1:
The patent applies different color adjustment strategies to different regions of the image. Water regions are protected from color correction to maintain their blue tint, while non-water regions (such as subjects and objects) are corrected to remove the blue cast. This local differentiation resolves the contradiction by preserving underwater authenticity in water areas while achieving color accuracy in other areas.
Solution Approach 2:
The image processing is segmented into identifying water regions and non-water regions, with separate color adjustment operations applied to each segment. This segmentation allows the system to selectively correct colors only where appropriate while preserving the characteristic blue water color, thus maintaining both color accuracy and underwater scene authenticity.
2Measurement precision
If standard white balancing is applied to remove the blue tint, then the overall color balance is improved, but the characteristic blue water color is also removed
Solution Approach 1:
The patent implements local quality by applying different processing rules to different image regions. Water regions are excluded from standard white balancing operations, while non-water regions receive full color correction. This approach achieves good color balance in the image while preserving the essential blue water color that indicates an underwater scene.
3Measurement precision
If color correction is applied to all pixels, then the tint is removed from all objects, but the water color is also altered
Solution Approach 1:
The patent applies local quality by implementing region-specific color correction. Water pixels are identified and protected from color adjustment, maintaining their authentic blue appearance, while non-water pixels are corrected for accurate colors. This resolves the contradiction by ensuring color accuracy where needed while preserving image authenticity in water regions.
Data Source
AI summary
A system and method that receives and edits image data of an underwater scene in a digital image in order to remove undesirable tints from objects in the scene. In some embodiments, colors near the color of the water itself are protected to leave the water looking blue. Removing undesirable tints without removing the tint of the water itself results in images with more realistic coloring of people and objects in the scene, without eliminating the color cues (e.g., blue water) that indicate that the image is a photograph of an underwater scene.