Image Processing Apparatus Saturation Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing methods fail to prevent saturation decrease and overcorrection during gradation correction, particularly in skin tone areas, leading to undesirable image quality.

Innovation Solution

The solution involves setting a new equal saturation line after gradation correction and determining a saturation correction amount based on this line to maintain proper saturation levels, using a 3DLUT system for efficient processing, and adjusting the equal saturation line according to the photographed scene to prevent overcorrection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If only the luminance in the pixel value included in the dark area in the image is increased in the luminance color difference space, then the brightness of the object is improved, but the saturation of the pixel value is reduced

Engineering Contradiction:
ImprovebrightnessVSAvoidsaturation
Core Design Contradiction:
Illumination intensityVSReliability

Solution Approach 1:

The patent changes the parameter control strategy by introducing saturation awareness into the luminance adjustment process. Instead of uniformly increasing luminance, the system calculates saturation values before and after luminance increase, and dynamically adjusts the luminance correction amount based on saturation changes. This ensures that luminance improvement does not come at the cost of saturation loss, resolving the contradiction between brightness enhancement and saturation preservation.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If the method of keeping the relative distance to the color space boundary is uniformly applied with respect to all hues and saturations, then the saturation is maintained in a general sense, but overcorrection occurs in specific areas such as skin color

Engineering Contradiction:
Improvesaturation maintenanceVSAvoidcolor accuracy
Core Design Contradiction:
ReliabilityVSManufacturing precision

Solution Approach 1:

The patent applies local quality by treating different hue regions differently. Instead of a uniform saturation maintenance approach, the system calculates saturation values specifically for each pixel's hue region and compares them against region-appropriate thresholds. This allows skin tones and other sensitive color regions to receive customized correction amounts, preventing overcorrection while maintaining saturation in other areas. The correction is locally adapted to each pixel's color characteristics rather than applying a global rule.

Inventive Principle:
Principle #3Local quality

3Productivity

If a predetermined amount is uniformly added to all components in the RGB space to increase luminance, then the brightness correction is simple and fast, but the relative differences among RGB components are reduced causing saturation decrease

Engineering Contradiction:
Improveprocessing speedVSAvoidsaturation
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent applies preliminary action by calculating the saturation value before luminance correction and using it to determine the appropriate correction amount. The system performs a preliminary saturation assessment, then uses this information to guide the luminance adjustment process. This preliminary saturation calculation allows the system to predict and prevent saturation loss before it occurs, rather than attempting to correct it afterward, maintaining both processing efficiency and saturation integrity.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2433418B1Image processing apparatus, method, and program-recorded storage medium
Publication Date: 2016.04.20 CANON KK
  • EP2433418B1 patent drawingFigure 1
  • EP2433418B1 patent drawingFigure 2
  • EP2433418B1 patent drawingFigure 3

AI summary

A gradation correcting curve to correct a gradation of an input image is generated. Gradation correction based on the gradation correcting curve is made to a boundary pixel of a color reproduction space at the same saturation as that of a target lattice point and at a hue of the target lattice point. An equal saturation line is set by using the boundary pixel of the color reproduction space after the gradation correction. A saturation correction amount of the target lattice point is decided based on the gradation correcting curve and the equal saturation line.