Handwriting Recognition via Curve Segmentation and Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current handwriting recognition methods are limited in recognizing cursive words due to complexity introduced by segmentation, with most methods relying on Neural Networks and Hidden Markov Models that struggle to efficiently process connected scripts.
Innovation Solution
A method for recognizing handwritten patterns involving curve segmentation, normalization, and matching measures that compensate for translation, angle, and length differences, using segmental and connective features, and an operator to determine matching templates, allowing for efficient comparison and recognition of cursive words.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Neural Networks and Hidden Markov Models are used for cursive word recognition, then recognition capability is improved, but device complexity increases
Solution Approach 1:
The handwritten pattern is divided into multiple segments by detecting segmentation points on curves. Each segment is processed independently through normalization and feature extraction, then combined through matching measures. This segmentation approach simplifies the overall recognition task compared to processing entire cursive words as single units, reducing device complexity while maintaining recognition capability.
2Measurement precision
If segmentation is performed on cursive words, then recognition accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent applies normalization transformations to segment parameters (scaling, translation, rotation) to make them invariant to writing variations. This parameter transformation approach simplifies the matching process by reducing the dimensionality of comparisons needed, improving recognition accuracy without proportionally increasing processing complexity.
Solution Approach 2:
Normalization and feature extraction are performed on segments before the actual matching process. This preliminary processing prepares segments in a standardized form, making subsequent matching operations more efficient and accurate, thereby improving recognition accuracy while controlling processing complexity through structured preprocessing.
3Productivity
If segmental and connective features are extracted and normalized, then matching efficiency is improved, but computational requirements increase
Solution Approach 1:
The patent extracts specific segmental features (local curve characteristics) and connective features (relationships between segments) from the handwritten pattern. By focusing on these key features rather than processing all raw data points, the method improves matching efficiency while reducing computational requirements through selective feature extraction.
Solution Approach 2:
The normalization scheme and matching measures are designed to handle multiple types of features (segmental and connective) using unified mathematical operations. This universal approach allows the same computational framework to process different feature types, improving matching efficiency across various feature combinations without proportionally increasing computational requirements.
Data Source
AI summary
A method for recognition of a handwritten pattern comprising one or more curves is presented. The method comprises a step of receiving sample data representing the handwritten pattern. The method further comprises a step of segmenting the handwritten pattern by detecting segmentation points on each curve, and by dividing the handwritten pattern into segments. Further, the method comprises a step of comparing the handwritten pattern to templates wherein the comparing comprises a step of normalizing said segments according to a scheme which is independent of the templates to which the segments are to be compared, and a step of determining matching measures for selecting at least one sequence of templates representing a recognintion candidate of the handwritten pattern.