Handwriting Recognition Using Segmentation Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for cursive handwriting recognition are complex and require significant processing power, often relying on large neural networks and dictionaries, which are slow and dependent on dictionary size, making them inefficient for real-time recognition.
Innovation Solution
A method that selects possible segmentation points in a handwritten pattern to segment and recognize symbols, comparing these segments to templates and a dictionary, using a Trie-graph for efficient comparison, thereby reducing computational efforts and eliminating the need for complex models like neural networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If neural networks and hidden markov models with integrated dictionaries are used for cursive handwriting recognition, then recognition accuracy is improved, but device complexity and processing time increase significantly
Solution Approach 1:
The handwritten pattern is segmented into multiple segments at selected segmentation points, creating a segmentation graph where each node represents a segmentation point and edges represent possible segments. This divides the complex recognition problem into smaller, manageable segments that can be processed independently and then combined, reducing overall system complexity while maintaining recognition accuracy.
Solution Approach 2:
Segmentation points are selected in advance based on predefined criteria (such as curvature thresholds or stroke characteristics) before the actual recognition process. This preliminary selection of segmentation points creates a structured framework that guides subsequent template matching, eliminating the need for complex real-time decision-making about where to segment the handwriting.
2Reliability
If large dictionaries are used to improve recognition results by evaluating probability of correct recognition, then recognition accuracy is improved, but processing speed decreases
Solution Approach 1:
The dictionary matching process is segmented and distributed across multiple nodes in the segmentation graph. Each node performs local template matching and stores segment candidates with associated measures independently. This parallel processing approach allows the system to evaluate multiple segmentation possibilities simultaneously without requiring sequential processing of the entire dictionary, thereby maintaining accuracy while improving processing speed.
Solution Approach 2:
Instead of performing exhaustive dictionary matching on the entire handwritten pattern, the system performs partial matching at each segmentation node, generating segment candidates that are then combined. This partial action approach processes only the necessary portions of the dictionary at each node, reducing overall processing time while still achieving accurate recognition through the combination of multiple partial results.
3Adaptability or versatility
If many possible segmentation points are considered for cursive handwriting recognition, then recognition completeness is improved, but computational effort increases
Solution Approach 1:
The segmentation graph structure divides the computational task into discrete nodes and edges, where each node represents a segmentation point and each edge represents a possible segment. This segmentation allows the system to efficiently traverse only relevant paths in the graph based on template matching results, avoiding exhaustive evaluation of all possible segmentations while maintaining recognition completeness.
Solution Approach 2:
Segmentation points are pre-selected based on objective criteria such as curvature thresholds, stroke direction changes, or pressure variations in the handwriting. This preliminary selection filters out unlikely segmentation points before template matching occurs, reducing the number of segments that need to be evaluated while ensuring that all plausible segmentation options are considered for accurate cursive handwriting recognition.
Data Source
AI summary
A method for determining at least one recognition candidate for a handwritten pattern comprises selecting possible segmentation points in the handwritten pattern for use in segmenting and recognizing the handwritten pattern. The method further may comprise comparing segments of the handwritten pattern to templates. The comparison may return segment candidates forming possible recognition results of the segments of the handwritten pattern. The method further comprises forming a representation of sequences of segment candidates, said representation comprising data blocks corresponding to segmentation points, wherein a data block comprises references to data blocks corresponding to subsequent segmentation points. The reference may comprise information of segment candidates. The method further may comprise comparing the representation of the sequences of segment candidates to a dictionary, finding sequences of segment candidates that correspond to allowed sequences of symbols in the dictionary, and returning at least one of these allowed sequences of symbols as a recognition candidate for the handwritten pattern.