Weighted Color Pixel Correction for Image Granularity

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image processing systems face challenges in achieving better granularity when determining whether to correct a pixel of interest based on the number or ratio of color pixels within a predetermined region, leading to uneven density and inferior image quality.

Innovation Solution

An information processing system that performs a weighting operation on color pixels surrounding a pixel of interest, assigning lower weights to pixels farther away, and corrects the pixel of interest if the sum of weights is lower than a threshold.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If the number or ratio of color pixels within a predetermined region is used to determine whether to correct a pixel of interest, then the determination process is simple, but the image granularity becomes inferior and uneven density occurs

Engineering Contradiction:

Improvedetermination process complexityVSAvoidimage granularity

Core Design Contradiction:

Device complexityVSManufacturing precision

Solution Approach 1:

The patent changes the parameter from simple pixel counting to weighted summation, where each color pixel contributes differently based on its distance from the pixel of interest. This transforms the determination criterion from a discrete count to a continuous weighted value, enabling more nuanced detection of isolated points while maintaining computational efficiency.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent applies different weights to color pixels based on their local position relative to the pixel of interest. Pixels closer to the center have higher weights, while those farther away have lower weights. This local differentiation allows the system to focus on the most relevant pixels for determining isolation, improving granularity without excessive complexity.

Inventive Principle:

Principle #3Local quality

2Productivity

If a threshold-based correction method is used for color pixels in a region, then the processing is efficient, but isolated points are not accurately detected leading to poor image quality

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidisolated point detection accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent transforms the detection parameter from a simple count of color pixels to a weighted sum where each pixel's contribution depends on its distance from the pixel of interest. This parameter change enables more accurate isolated point detection while maintaining a straightforward threshold comparison process for efficient processing.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical counting approach with a weighted evaluation system that substitutes simple arithmetic with a more sophisticated but still computationally efficient mathematical model. This substitution improves detection accuracy without proportionally increasing processing complexity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20250080684A1Information processing system, non-transitory computer readable medium and information processing method

Publication Date: 2025.03.06 FUJIFILM BUSINESS INNOVATION CORP

AI summary

An information processing system includes a processor configured to: perform a weighting operation to weight, about image data on white pixels and color pixels, color pixels surrounding a pixel of interest within a predetermined region in a manner such that a lower weight is attached to each of the color pixels as the color pixel is farther from the pixel of interest; and correct the pixel of interest if a sum of weights of the color pixels surrounding the pixel of interest is lower than a threshold.