Character Recognition Using Segmented Template Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing character template recognition method in image processing achieves poor recognition effects due to its simplicity and lack of accuracy in differentiating between characters.
Innovation Solution
A character information recognition method that involves creating a character template library with both coincidence-permitted and coincidence-restricted regions, where the coincidence rate is calculated by comparing target character images with template images, using a formula that weights coinciding and non-coinciding pixels to determine the accuracy of recognition, and employing scanning methods to reduce errors in alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If character template setting method is used, then the algorithm complexity is reduced, but the recognition accuracy deteriorates
Solution Approach 1:
The character template is segmented into two functional regions: coincidence-permitted region and coincidence-restricted region. The coincidence-permitted region allows flexible matching for robustness, while the coincidence-restricted region enforces precise differentiation between similar characters. This segmentation enables the system to achieve both simplicity in template construction and accuracy in character differentiation.
Solution Approach 2:
Different regions of the character template are assigned different matching qualities. The coincidence-permitted region uses lenient matching criteria to handle variations in font and rendering, while the coincidence-restricted region uses strict matching criteria to differentiate between characters with similar overall shapes. This local differentiation of matching quality resolves the contradiction between simplicity and accuracy.
2Measurement precision
If neural network algorithm is used, then the recognition accuracy is improved, but the algorithm complexity and data collection requirements worsen
Solution Approach 1:
Instead of using complex neural networks to learn character representations, the invention creates a library of standardized character template images that serve as copies of expected character appearances. The recognition process then involves comparing the input character image against these pre-prepared templates using simple similarity metrics, avoiding the need for complex training algorithms while maintaining high recognition accuracy.
Solution Approach 2:
The character templates are prepared in advance during the preprocessing stage, where standard character images are generated and stored in the template library. This preliminary action eliminates the need for complex algorithms during the actual recognition process, as the system only needs to compare the input against the pre-computed templates, significantly reducing algorithmic complexity.
3Productivity
If simple character template comparison is used, then the processing speed is improved, but the differentiation capability between similar characters worsens
Solution Approach 1:
The template comparison is segmented into two stages: first comparing the coincidence-permitted region for rapid initial matching, then refining the comparison using the coincidence-restricted region for precise differentiation. This segmented approach maintains fast processing speeds while enhancing the ability to differentiate between similar characters like '0' and 'O' or '1' and 'l'.
Solution Approach 2:
The coincidence-restricted region is designed to contain specific local features that are critical for differentiation between similar characters. By focusing the strict comparison on these key regions rather than the entire character template, the system achieves fast processing with high differentiation capability, as the restricted region quickly identifies whether two characters are truly the same or merely similar.
Data Source
AI summary
The present invention relates to a character information recognition method based on image processing. The method comprises: collecting images to obtain a target character image; then sequentially comparing the target character image with character template images in a character template library to find a maximum of a coincidence area of the character in the target character image with the character templates in the character template images; and when the coincidence area meets a preset condition, determining the target character to be recognized as the character in the corresponding character template image. The character templates are designed to include not only a coincidence-permitted region but also a coincidence-restricted region. The coincidence-restricted region is set, so that the direct comparing and matching of the character templates can be more accurately carried out, thereby improving the recognition speed.