OCR-Specific Image Correction for Color and Seal Interference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OCR processing techniques struggle with accurately extracting text information from scanned images containing overlapping color content or seal impressions, leading to reduced recognition rates due to varying image quality requirements among different OCR software.
Innovation Solution
An image processing apparatus and method that adjusts image correction based on OCR software information, including color removal, seal impression removal, and resolution settings to optimize image quality for specific OCR software, enhancing recognition accuracy without significantly altering the original image appearance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image correction is performed to improve OCR recognition accuracy, then character recognition accuracy is improved, but image quality and appearance may be degraded
Solution Approach 1:
The patent applies parameter changes by adjusting image correction parameters based on the type of OCR software detected. Different correction parameters are selected according to whether the software requires color removal, seal impression removal, or specific resolution settings, thereby optimizing the balance between recognition accuracy and image quality preservation for each software type.
Solution Approach 2:
The patent implements dynamics by making the image correction process adaptive and variable. The correction parameters are not fixed but dynamically adjusted based on real-time detection of OCR software requirements, allowing the system to optimize image processing according to different software needs while maintaining both accuracy and quality.
2Reliability
If image correction is tailored to specific OCR software requirements, then OCR processing accuracy is improved, but device complexity increases
Solution Approach 1:
The patent applies self-service by having the system automatically detect and identify the type of OCR software being used, then autonomously select and apply the appropriate image correction parameters without manual intervention. This automation reduces the operational complexity while maintaining high processing accuracy through software-type adapted corrections.
Solution Approach 2:
The patent implements preliminary action by pre-defining multiple sets of correction parameters corresponding to different OCR software types before processing begins. The system detects the software type and retrieves the pre-configured parameters, avoiding complex real-time calculations and reducing processing complexity while ensuring accuracy for each software type.
3Measurement precision
If color removal and seal impression removal are performed, then character recognition is improved, but original image appearance is altered
Solution Approach 1:
The patent applies local quality by selectively applying different correction treatments to different regions and elements of the image based on their function. Color removal is applied selectively to color content that interferes with characters, while seal impressions are removed only when detected and when required by the OCR software, thereby maintaining character recognition accuracy while preserving original image appearance where possible.
Data Source
AI summary
An image processing apparatus includes an imager and circuitry. The imager captures an image. The circuitry sets Optical Character Recognition (OCR) software information corresponding to OCR software used for subsequent OCR processing, performs image correction set according to the set OCR software information on the image captured by the imager to obtain a corrected image, and outputs the corrected image.