Offline Character Recognition Segmentation for Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Offline handwriting recognition systems face challenges in accuracy due to the inability to edit or correct recognized text in real-time and the variability in handwriting styles and languages, leading to accumulated errors.
Innovation Solution
A character recognition method that receives an image of handwritten text, segments it into strokes, determines parameters for each stroke, and compares these parameters with stored sets to identify matching strokes, allowing for accurate recognition across different handwriting styles and languages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If offline handwriting recognition is implemented to recognize characters from images, then the system can process static handwritten text, but the recognized text cannot be edited or corrected until the entire text is recognized, leading to accumulated errors
Solution Approach 1:
The patent segments the handwritten text recognition process into individual character or word units. Each segment is recognized and presented to the user separately rather than waiting for complete text recognition, allowing users to correct errors incrementally as they appear rather than after the entire text is processed.
Solution Approach 2:
The system implements immediate feedback by presenting recognized text segments to users right after recognition, before processing continues. This allows users to provide correction feedback on individual characters or words immediately, preventing error accumulation across the entire text.
2Productivity
If the system recognizes entire handwritten text before presentation, then complete text can be processed, but users cannot provide feedback for correction on individual characters or words
Solution Approach 1:
The patent divides the handwritten text into recognizable segments (characters or words) that can be processed and presented individually. This segmentation enables the system to maintain batch processing efficiency while allowing incremental user corrections, as each segment can be independently verified and corrected.
Solution Approach 2:
The system performs preliminary recognition on individual text segments before presenting them to users. This preliminary action allows the system to prepare recognition results in advance while still enabling user correction, as the segments are ready for review before being finalized.
3Adaptability or versatility
If the system is designed to handle varied handwriting styles and languages, then it can recognize diverse text, but recognition accuracy decreases due to character similarities across different languages and styles
Solution Approach 1:
The patent segments text recognition into smaller units (characters or words) that are processed individually. This segmentation allows the system to apply language-specific or style-specific recognition models to each segment, improving precision while maintaining versatility across different handwriting styles and languages.
Solution Approach 2:
The system applies different recognition characteristics or parameters to different segments based on their local properties (language, handwriting style). This local quality approach allows the system to optimize recognition precision for each segment while maintaining overall adaptability to diverse handwriting variations.
Data Source
AI summary
A method, non-transitory computer readable medium and character recognition device for character recognition includes a character recognition device for receiving an image representing a character comprising one or more strokes. A set of first parameters associated with each of the one or more first strokes is determined. Each of the one or more first strokes is compared with the associated set of first parameters with a plurality of stored sets of second parameters, where each of the plurality of stored second strokes is associated with a stored set of second parameters. A second stroke is identified from among the plurality of stored second strokes corresponding to each of the one or more first strokes based on the comparison. The character is identified based on the identified one or more second strokes.