Character Recognition Image Grouping for Precise Region Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing character recognition technologies, such as CRAFT and Star-Net, face accuracy issues due to imprecise clipping of character regions, leading to reduced recognition performance, especially when characters are not uniformly spaced or distorted.
Innovation Solution
A method involving probability image analysis to estimate character positions, cluster characters into groups, and recognize characters from segmented images, ensuring accurate character region extraction and recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If character string region is clipped based on probability images, then character recognition can be performed, but recognition accuracy is reduced due to imprecise clipping
Solution Approach 1:
The patent segments the character string region into multiple candidate regions by dividing the clipped area into several sections. Each section is evaluated independently to determine the most appropriate character string region, thereby improving clipping precision while maintaining recognition accuracy through multi-candidate evaluation
Solution Approach 2:
The patent changes the evaluation parameter from simple probability thresholding to a composite scoring system that considers multiple factors including character density, spacing uniformity, and distortion metrics. This parameter transformation enables more accurate region selection that balances clipping precision with recognition reliability
2Reliability
If character string region includes more characters, then clustering becomes more complex, but recognition coverage improves
Solution Approach 1:
The patent applies segmentation by dividing the character string region into multiple candidate sections and evaluating each independently. This reduces clustering complexity by working with smaller, more manageable subsets while maintaining comprehensive recognition coverage through evaluation of multiple candidates
Solution Approach 2:
The patent uses partial action by selecting only the most promising candidate regions for detailed evaluation rather than processing all possible character combinations. This approach achieves sufficient recognition coverage without the full computational burden of exhaustive clustering analysis
Data Source
AI summary
The information processing device acquires a probability image representing a probability of an existence of a character in each of pixels included in a target image including a plurality of characters based on the target image, estimates positions of respective character images included in the target image based on the acquired probability image, classifies the plurality of character images into a plurality of groups based on the estimated positions, acquires a plurality of recognition target images which is generated so as to correspond to the plurality of groups, and includes the plurality of character images respectively belonging to the corresponding groups, and recognizes the plurality of characters from each of the recognition target images.