Centerline OCR Template Matching for Faster Character Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing optical character recognition (OCR) methods are time-consuming due to the need for extensive pixel analysis and template matching, particularly when using outline templates, which do not efficiently utilize a limited set of points for scoring character matches.
Innovation Solution
The method focuses on character centerlines, using depleted centerline templates with reduced points (about 20) and incorporates a primary and secondary analysis to assess match scores, including proximity to character boundaries, significantly reducing processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If outline templates are used for OCR decoding, then character recognition accuracy is maintained, but processing time increases significantly
Solution Approach 1:
The invention extracts only the essential centerline points from complete character templates, creating depleted centerline templates that retain recognition accuracy while reducing the number of points to approximately 20 per character. This extraction principle directly resolves the contradiction by removing unnecessary template points that contribute to processing time without significantly impacting recognition accuracy.
Solution Approach 2:
The invention segments the character template into two distinct components: the depleted centerline template (approximately 20 points) and the boundary information. This segmentation allows the system to perform rapid initial matching using only the centerline points, then selectively apply boundary verification only when needed, thereby reducing overall processing time while maintaining accuracy.
2Measurement precision
If pixel analysis is performed for each template position, then accurate character matching is achieved, but computational complexity increases
Solution Approach 1:
The invention applies partial action by performing complete pixel analysis only for the top-ranked template matches identified through rapid centerline comparison. For lower-ranked matches, the system uses simplified scoring based on fewer points. This selective application of analysis depth reduces computational complexity while maintaining accurate character matching for the most likely candidates.
3Reliability
If complete templates are used for matching, then comprehensive character coverage is achieved, but the number of points to be processed increases
Solution Approach 1:
The invention extracts only the critical centerline points from complete character templates, creating depleted centerline templates with approximately 20 points per character compared to the hundreds of points in complete templates. This extraction maintains reliability by preserving the essential structural information needed for accurate character identification while dramatically reducing the quantity of points to be processed.
4Reliability
If template matching is performed at every candidate location, then no character is missed, but processing time increases
Solution Approach 1:
The invention performs preliminary rapid centerline comparison at all candidate locations to identify and rank potential matches before conducting more time-consuming detailed analysis. This preliminary action filters out obviously incorrect matches early, allowing the system to maintain complete character detection by analyzing only the most promising candidates in detail, thereby reducing overall processing time.
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A method for template matching can include iteratively selecting a template set of points to project over a centerline of a candidate symbol; conducting a template matching analysis; assigning a score to each template set; and selecting a template set with a highest assigned score. For example, the score can depend on proximity of the template points to a center and/or boundaries of a principal tracing path of the symbol. Additionally, one or more template sets having a top rank can be selected for a secondary analysis of proximity of the template points to a boundary of a printing of the symbol. The method can further include using the template with the highest score to interpret the candidate symbol.