Binary Image Generation for Form Seal Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing character recognition systems struggle to accurately extract character information from form images when characters overlap with seals, as they often mistakenly identify characters with similar color, size, and density as seals and remove them, leading to incorrect extraction.
Innovation Solution
An information processing system determines candidate areas in an input image based on specific color pixels, differentiates processing targets, and generates a binary image by converting specific color pixels into white pixels, thereby preventing mistaken identification and removal of characters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pixels of a specific color (e.g., red) are extracted and converted into white pixels to remove seals, then seal removal effectiveness is improved, but character information accuracy deteriorates due to mistaken identification of characters with similar color properties
Solution Approach 1:
The patent changes multiple parameters simultaneously including color (specific color extraction), size (area threshold), and density (pixel density threshold) to differentiate seals from characters. By setting specific threshold values for area and pixel density, the system can distinguish seals from characters even when they share similar color properties, thus resolving the contradiction between effective seal removal and character preservation
Solution Approach 2:
The system performs feedback by detecting the area and pixel density of extracted color regions, comparing them against threshold values, and adjusting the processing accordingly. This feedback mechanism allows the system to identify whether a detected region is a seal or a character based on multiple parameters, preventing mistaken identification while maintaining effective seal removal
2Difficulty of detecting and measuring
If a circumscribed rectangle is extracted based on color pixel density and size to identify seal areas, then seal detection capability is improved, but false positive rate increases causing character strings to be erroneously removed
Solution Approach 1:
The patent introduces multiple parameters (area threshold, pixel density threshold, and color specificity) to improve seal detection accuracy. By requiring simultaneous satisfaction of multiple parameter conditions rather than relying on a single parameter, the system reduces false positives while maintaining high seal detection capability
Solution Approach 2:
The patent transitions from single-parameter detection to multi-dimensional detection by considering color, area, and pixel density simultaneously. This dimensional expansion allows the system to distinguish seals from characters more accurately, reducing false positives while maintaining detection capability
Data Source
AI summary
In the present disclosure, a candidate area is determined based on a pixel having a specific color included in an input image, and an area is determined to be a processing target from the candidate area based on a pixel having a predetermined color different from the specific color included in the candidate area. Further, a second binary image in which a pixel corresponding to the pixel having the specific color is converted into a white pixel is generated by converting, in a first binary image obtained by the input image being binarized, a pixel that is included in the area determined to be the processing target and corresponds to the pixel having the specific color, into a white pixel.