Ink Parser for Hand-Drawn Chart Recognition and Editing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer systems are unable to effectively recognize and edit hand-drawn diagrams, charts, lists, and tables in ink input, as existing algorithms are sensitive to stroke order and number, and fail to group strokes correctly, limiting the editing capabilities of such objects.
Innovation Solution
A system and method that includes an ink parser coupled with an ink editing user interface, featuring detectors and recognizers for charts, lists, and tables, allowing for the recognition and editing of ink objects independently of stroke order and number, with editors for specific types of objects and a mode switcher between inking and editing modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If incremental recognition algorithms are used to recognize simple geometric shapes, then shape recognition capability is improved, but the system fails to be robust because it relies on stroke order and assumes a particular number of strokes
Solution Approach 1:
The patent changes the parameters of stroke recognition from fixed (specific number of strokes, specific order) to variable (any number of strokes, any order). The system uses clustering algorithms to group strokes into shapes without requiring a predetermined number of strokes or specific drawing sequence, making the recognition robust to variations in user input patterns.
Solution Approach 2:
The patent introduces dynamic grouping where strokes are clustered into shapes based on spatial proximity and geometric characteristics rather than static rules. The clustering process dynamically determines which strokes belong together based on their relative positions and orientations, allowing the system to adapt to different drawing styles and stroke counts.
2Device complexity
If incremental algorithms assume a particular number of strokes, then recognition process is simplified, but the system cannot accommodate multi-stroke shapes or hand-drawn objects
Solution Approach 1:
The patent creates a universal recognition system that handles both simple geometric shapes and complex hand-drawn objects using the same clustering-based approach. The system can recognize circles, rectangles, and more complex diagrams and charts without requiring different algorithms, achieving multi-functionality through a single robust framework.
Solution Approach 2:
The system changes from fixed-parameter recognition (specific stroke counts) to variable-parameter recognition where the number of strokes per shape is dynamically determined through clustering. This allows the same algorithm to handle shapes with 2 strokes, 5 strokes, or any other number of strokes.
3Measurement precision
If the system requires specific stroke order for recognition, then recognition accuracy for simple shapes is improved, but the system becomes sensitive to input variations and less robust
Solution Approach 1:
The patent implements dynamic stroke grouping where the system determines shape membership based on spatial relationships rather than temporal sequence. Strokes are clustered based on their geometric proximity and orientation, allowing recognition regardless of the order in which strokes were drawn, making the system reliable across different input variations.
4Ease of operation
If menu-based application programs are used to create drawings, then editing capability is improved, but the ability to directly recognize and edit hand-drawn ink objects is lost
Solution Approach 1:
The patent enables hand-drawn ink objects to be automatically recognized and made editable without requiring manual intervention through menu-based programs. The system self-services by automatically detecting shapes, clusters, and diagrams in the ink input and converting them into editable structures, bridging the gap between freehand drawing and structured editing.
Data Source
AI summary
A system and method for editing ink objects recognized in ink input is provided. An ink parser may recognize an ink object in ink input and then an ink editing user interface may edit the ink object recognized by the ink parser. The ink parser may include a chart detector, shape recognizer, and various ink object recognizers such as a chart recognizer, a list detector and a table detector. The various ink object recognizers may recognize particular ink objects. The ink editing user interface may edit the ink object recognized by the ink parser. The ink editing user interface may include a chart editor, list editor, table editor, mode switcher, and a visualizer. The mode switcher may switch the ink editing system between inking mode and ink editing mode.