Image Processing Device Dynamic Grading for Text and Photo Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies fail to effectively output images with an appropriate number of gradations, particularly when converting documents to binary or multivalued formats, leading to issues like character crushing and quality deterioration in photographic regions.
Innovation Solution
An image processing device that includes determiners for character crushing and photographic region rate assessment, allowing conversion into monochrome N-gradation image data based on these determinations to optimize output quality and file size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If image data is converted to binary format for compression, then file size is reduced, but character crushing occurs and image quality deteriorates
Solution Approach 1:
The patent dynamically changes the number of gradations (parameter) based on image content analysis. For text-heavy images, it uses fewer gradations (binary or 4-level) to reduce file size, while for photographic images, it increases gradations to prevent quality deterioration. This parameter adaptation resolves the contradiction between compression and quality preservation.
Solution Approach 2:
The patent segments the image into different regions (text regions and photographic regions) and applies different gradation levels to each segment. Text regions use binary or low-gradation formats for compression, while photographic regions use higher gradations to maintain quality, thus resolving the contradiction at the regional level.
2Quantity of substance
If image data is converted to monochrome multivalued format, then file size is reduced compared to grayscale, but character crushing occurs in text regions
Solution Approach 1:
The patent applies different quality levels to different parts of the image. Text regions are processed with binary conversion and character crushing prevention algorithms, while photographic regions use monochrome multivalued conversion. This local differentiation resolves the contradiction by optimizing each region independently.
Solution Approach 2:
The patent dynamically selects the appropriate conversion method (binary or monochrome multivalued) based on real-time analysis of image content characteristics. The system adapts the processing parameters according to whether the image contains text, photographs, or mixed content, resolving the contradiction through dynamic decision-making.
3Measurement precision
If automatic color selection processing is performed on separated image regions, then processing accuracy is improved, but only single determination results are used without considering multiple factors
Solution Approach 1:
The patent merges multiple determination results (character crushing determination, photographic region determination, text region determination) into a unified processing decision. By combining these determinations, the system achieves higher accuracy in selecting the appropriate conversion method while managing complexity through integrated processing logic.
Solution Approach 2:
The patent uses feedback from multiple determination processes to adjust the conversion parameters. The results of character crushing determination, photographic region detection, and text region analysis feed back into the decision-making process, allowing the system to optimize the conversion method based on comprehensive image characteristics.
Data Source
AI summary
An image processing device includes an image reader that reads an image, a first determiner that determines whether character crushing occurs when the image is binarized, a second determiner that determines a rate of a photographic region in the image, and a controller that performs conversion into monochrome N-gradation image data based on determination results of the first and second determiners.