Optical Character Recognition Using Fuzzy Pattern Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional optical character recognition (OCR) techniques, especially for handwritten text, face challenges due to variations in handwriting styles, resulting in low accuracy when encountering input that differs from the training data, as they rely heavily on machine learning models that require large datasets and struggle with uncertainty in predictions.
Innovation Solution
A system performs a set of transformations on input images, including scaling, contrast changes, and noise addition, to generate candidate text predictions, which are then used to build a regular expression based on confidence scores for exact or wildcard matches, allowing for accurate text recognition by matching against a database.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If machine learning models are used for OCR of handwritten text, then text recognition capability is improved, but accuracy deteriorates when input differs from training data
Solution Approach 1:
The system performs image transformations (scaling, rotation, contrast changes, noise addition) on the input image before OCR recognition to pre-process and normalize various handwriting variations, making the input more consistent with training data distributions and improving prediction accuracy
2Productivity
If conventional OCR techniques are used, then processing speed is maintained, but accuracy deteriorates for handwritten text with varying styles
Solution Approach 1:
The system changes multiple image parameters simultaneously (scale, rotation angle, contrast, noise level) to generate transformed versions of the input image, capturing the variability in handwriting styles and improving recognition accuracy without significantly impacting processing speed
3Measurement precision
If multiple image transformations are performed to account for handwriting variations, then recognition accuracy is improved, but system complexity increases
Solution Approach 1:
The system segments the complexity by dividing the transformation process into distinct, independent operations (scaling, rotation, contrast adjustment, noise addition) that can be applied separately and combined, making the overall complex process more manageable and controllable
Data Source
AI summary
A system recognizes text in an input image. The system provides the input image to one or more optical character recognition (OCR) models to obtain predicted texts. The system determines a set of candidate text predictions by performing text recognition on each transformed image of the set of transformed images. The system generates a regular expression based on the predicted characters of the candidate text predictions and confidence score corresponding to each predicted character. The system matches the regular expression against text values in a database. The system selects one or more text values from the database based on the matching and returns the one or more text values as results of recognition of text of the input image.