Digital Ink Text Line Extraction for Non-Chronological Handwriting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computing devices face challenges in accurately extracting text lines and blocks from user input handwriting, particularly when the handwriting is non-chronological, leading to errors in text recognition due to diacritics, punctuation marks, and delayed strokes, which disrupt the temporal order and increase the complexity of text line extraction.
Innovation Solution
A method involving slicing the display area into transversally extending strips, ordering strokes into timely and spatially ordered lists, performing neural net analysis to determine text line associations, and defining lines based on probability scores, ensuring accurate text line extraction even in free handwriting formats.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If free handwriting mode is used to allow complete user freedom, then ease of operation is improved, but text line extraction accuracy deteriorates due to non-chronological strokes
Solution Approach 1:
The patent segments the handwriting recognition process into multiple passes: first pass processes strokes in temporal order to establish initial text lines, second pass processes remaining strokes to correct and refine line assignments. This segmentation allows the system to maintain both free handwriting mode and accurate text line extraction by handling different stroke types in appropriate sequences.
Solution Approach 2:
The patent performs preliminary text line extraction on chronologically ordered strokes before processing delayed strokes. This preliminary action establishes a baseline structure that can later be refined, allowing the system to maintain accuracy even when users write in non-chronological order.
2Adaptability or versatility
If delayed strokes are processed to accommodate non-chronological handwriting, then adaptability is improved, but text line extraction complexity increases
Solution Approach 1:
The patent implements a dynamic two-pass processing system that adapts to the actual handwriting pattern. The first pass handles chronological strokes efficiently, while the second pass selectively processes only the delayed strokes that require correction. This dynamic approach maintains adaptability without permanently increasing system complexity.
Solution Approach 2:
The system uses feedback from the first pass to identify which strokes were misplaced and require reprocessing. The second pass uses this feedback information to focus computational resources only on the problematic delayed strokes, rather than reprocessing all strokes, thereby managing complexity effectively.
3Reliability
If multiple processing passes are used to improve accuracy, then text line extraction reliability is improved, but processing time increases
Solution Approach 1:
The patent applies partial action by processing only the necessary subset of strokes in the second pass - specifically, only the delayed strokes that were not correctly assigned in the first pass. This avoids excessive processing of all strokes and minimizes time loss while maintaining high reliability.
Solution Approach 2:
By performing preliminary processing of chronological strokes in the first pass, the system establishes correct text line assignments early, so that the second pass only needs to handle corrections for delayed strokes. This preliminary action reduces the overall processing burden.
Data Source
AI summary
The invention concerns a method implemented by a device for displaying strokes of digital ink in a display area and for performing text line extraction to extract text lines from the strokes. In particular, the text line extraction may involve slicing the display area into strips, ordering for each strip the strokes into ordered lists which form collectively a first set of ordered lists, forming for each strip a second set of ordered lists by filtering out from the ordered lists of the first set strokes which are below a given size threshold, and performing a neural net analysis based on said first and second sets to determine for each stroke a respective text line to which it belongs.