Color-Reduced Image Processing for Clear Scanned Characters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing color reduction methods result in blurred characters in scanned documents with luminance variations, leading to low object reproducibility.
Innovation Solution
An image processing apparatus that generates chroma and brightness binary images to determine representative colors, converting pixels based on their chroma and brightness attributes, and combines these to create a color-reduced image with high object reproducibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional color reduction processing is applied to scanned documents with luminance variations, then data size is reduced, but character clarity deteriorates due to blurring
Solution Approach 1:
The patent segments the color reduction process into two distinct stages: first performing color reduction on the entire image, then performing secondary color reduction only on background regions. This segmentation allows different processing strategies to be applied to different parts of the image, preserving character clarity while achieving overall data size reduction.
Solution Approach 2:
The patent applies different color reduction qualities to different regions: high-quality processing is applied to character regions to maintain clarity, while aggressive compression is applied to background regions where quality is less critical. This local quality approach resolves the contradiction between overall compression and local character preservation.
2Productivity
If aggressive color reduction is applied to achieve high compression, then storage efficiency improves, but object reproducibility deteriorates
Solution Approach 1:
The patent performs preliminary color reduction on the entire image before performing secondary color reduction on background regions. This preliminary action establishes a baseline compressed version, and the secondary processing selectively enhances critical regions, achieving both compression efficiency and object reproducibility.
Solution Approach 2:
The patent uses an intermediary classification process that identifies and separates background regions from character regions. This intermediary step enables differential processing where background regions undergo additional color reduction while character regions maintain higher quality, balancing compression efficiency with reproducibility.
3Device complexity
If single-stage color reduction is applied to reduce processing complexity, then device complexity decreases, but character preservation worsens due to uniform processing
Solution Approach 1:
The patent segments the processing into two stages with different complexity levels: a simple first color reduction pass followed by a more sophisticated second pass on background regions only. This segmentation manages overall complexity while improving character preservation through targeted processing.
Solution Approach 2:
The patent applies partial action by performing the second color reduction only on background regions rather than the entire image. This reduces the amount of processing required compared to uniform high-quality processing, while still achieving good character preservation through the selective approach.
Data Source
AI summary
An image processing apparatus capable of generating a color-reduced image having high object reproducibility. The image processing apparatus: determines a first representative color(s)/a second representative color/a third representative color, based on an RGB value of a pixel determined to have a high chroma attribute/a low chroma attribute/a low brightness attribute based on a chroma binary image/on a chroma binary image/on a brightness binary image, among pixels constituting an acquired image; generates a converted image by converting the RGB value of the pixel having the high chroma attribute/low chroma attribute into an RGB value indicating the first representative color/second representative color; generates a color-reduced image by converting the RGB value of the pixel determined to have the low brightness attribute among pixels constituting the converted image into an RGB value indicating the third representative color.