Image Compensation for Low Contrast Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image compensation techniques struggle to effectively remove distortion in scanned images, particularly in books, due to limitations in distinguishing foreground and background values, especially when the object is not in consistent colors, leading to difficulties in addressing the 'folding area' distortion caused by varying distances from the scan plane.
Innovation Solution
An image forming method and apparatus that generates profiles from pixel values in marginal areas to determine background values and correct image brightness, converting RGB values to brightness and chrominance for accurate compensation, allowing for effective removal of distortion regardless of consistent foreground and background colors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If histogram analysis is used to extract parameters for image compensation, then the folding area distortion can be removed, but the method fails when the image does not have consistent foreground and background colors
Solution Approach 1:
The patent divides the image into multiple horizontal rows and extracts pixel values from specific positions in each row to create a profile. This segmentation approach allows the method to work with images of varying color compositions by analyzing local characteristics rather than relying on global histogram analysis that assumes consistent foreground and background colors.
Solution Approach 2:
The patent changes the approach from using histogram-based parameter extraction to using profile-based parameter extraction. By converting RGB values to brightness values and analyzing the profile of pixel intensities across the image, the method adapts to images with inconsistent color schemes while maintaining compensation accuracy.
2Reliability
If conventional compensation methods are used, then images with consistent colors can be compensated, but images with varying colors (such as those containing pictures) cannot be properly distinguished for compensation
Solution Approach 1:
The patent segments the image into multiple rows and extracts profiles from specific positions, allowing it to handle images with varying content types (text, pictures, mixed content) by analyzing local brightness characteristics rather than relying on global color consistency assumptions.
Solution Approach 2:
The patent transforms the compensation approach by changing from color-based histogram analysis to brightness-based profile analysis. This parameter change enables reliable distinction between foreground and background even in images containing pictures with varying colors, thereby improving both reliability and adaptability.
3Manufacturing precision
If section-by-section compensation is performed based on histogram analysis, then the folding area can be corrected, but the method requires consistent foreground and background colors which limits its application
Solution Approach 1:
The patent segments the image into rows and extracts profiles from specific positions in each row, creating a simplified representation that captures the essential brightness variation pattern. This segmentation enables accurate distortion correction while simplifying the overall method by avoiding complex histogram analysis of the entire image.
Data Source
AI summary
To compensate an image, a profile thereof is obtained from pixel values in a marginal area of the image. A background value of the image determined from the profile and correction to the image is performed in accordance with the background value and the profile.