OCR Trigraph Resolution for Ambiguous Character Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical character recognition (OCR) systems face accuracy and processing time challenges when dealing with ambiguous characters, which can lead to delays in item delivery and reduced efficiency in high-speed sorting operations.
Innovation Solution
The use of trigraphs and xgraphs to resolve ambiguous characters by substituting possible characters, comparing these combinations to a reference set, and ranking solutions based on likelihood, either through frequency of occurrence or contextual analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCR systems use traditional character recognition methods, then processing speed is maintained, but accuracy decreases when encountering ambiguous characters
Solution Approach 1:
The system performs preliminary actions by pre-defining multiple possible character interpretations for ambiguous visual patterns before final recognition. When an ambiguous character is encountered, the system has already prepared alternative character assignments based on contextual analysis, allowing rapid selection without time-consuming real-time analysis.
Solution Approach 2:
The recognition process is segmented into multiple independent stages: initial character detection, ambiguity identification, contextual analysis, and final character assignment. This segmentation allows the system to process ambiguous characters through specialized sub-routines that maintain overall processing speed while improving accuracy.
2Measurement precision
If OCR systems create multiple alternate possibilities for ambiguous characters, then accuracy improves, but processing time increases
Solution Approach 1:
The system applies partial action by generating and evaluating only the necessary number of alternate character possibilities based on contextual constraints. Rather than exhaustively testing all possible interpretations, the system uses contextual information to limit the search space to the most likely candidates, reducing processing time while maintaining high accuracy.
Solution Approach 2:
Contextual feedback from surrounding characters and document structure is continuously used to refine and eliminate unlikely character interpretations. This feedback mechanism allows the system to rapidly converge on the correct character assignment without needing to evaluate all possibilities, reducing processing time.
3Measurement precision
If OCR systems shunt items with ambiguous characters to manual sorting, then processing accuracy improves, but productivity decreases
Solution Approach 1:
The system enables self-service by automatically resolving ambiguous characters through contextual analysis and alternative character generation. Items with previously problematic ambiguous characters are now processed autonomously by the automated system, eliminating the need for manual intervention and maintaining high throughput.
Solution Approach 2:
The system changes the operational parameters of automated sorting by introducing contextual analysis and multi-hypothesis evaluation. These parameter changes enable the automated system to handle ambiguous characters that would traditionally require manual sorting, thereby maintaining both accuracy and productivity.
Data Source
AI summary
Systems, methods, and software consistent with the present invention use combinations of two or more letters (xgraphs) or characters, such as three letter combinations (trigraphs), to aid in deciphering ambiguous characters in OCR systems. Embodiments consistent with the invention compare possible xgraphs constructed by substitution for the ambiguous OCR-read character(s) to a reference set of xgraphs. The reference set of xgraphs contains predetermined valid xgraphs, which may be ranked by preference. If a possible xgraph matches a reference xgraph, then the ambiguous character(s) are resolved in favor of the characters in the reference xgraph. Embodiments may also use the context of the ambiguous character string to aid in resolving the ambiguity where more than one possible solution exists.