Fuzzy Regular Expressions for Noisy Image Content Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Extracting content from documents stored as images is an error-prone process, particularly when analyzing images that contain words, numbers, and symbols, as existing methods struggle to accurately identify and process the content.

Innovation Solution

The system employs fuzzy regular expressions to process OCR results, generating data models and observation graphs to match content patterns, and applies format modifications to the extracted content, utilizing a document processing system with OCR, fuzzy matching, and format modification components.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional content extraction methods are used on image data, then the process can be performed, but the accuracy and reliability of content identification deteriorates

Engineering Contradiction:

Improvecontent extraction accuracyVSAvoidcontent identification reliability

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The system changes the parameter of pattern matching from exact matching to fuzzy matching using regular expressions. This allows the system to tolerate variations in OCR output (such as different fonts, resolutions, or scanning quality) and still accurately identify content patterns, thereby improving both accuracy and reliability of content extraction from image data

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system incorporates feedback mechanisms where OCR results are processed through pattern matching algorithms that can correct or adjust extracted content based on contextual information. This feedback loop allows continuous improvement of extraction accuracy by comparing and validating results against expected patterns

Inventive Principle:

Principle #23Feedback

2Measurement precision

If fuzzy regular expressions and pattern matching are applied to OCR results, then content identification precision is improved, but processing complexity increases

Engineering Contradiction:

Improvecontent identification precisionVSAvoidprocessing system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The processing system is segmented into distinct functional components: image acquisition module, OCR processing module, pattern matching module, and content extraction module. Each component handles a specific task independently, making the overall complex processing pipeline more manageable and maintainable while achieving high precision through coordinated operation of these modular segments

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS10373006B2Method and system for extracting alphanumeric content from noisy image data

Publication Date: 2019.08.06 OPEN TEXT CORP

AI summary

In general, embodiments of the technology relate to extracting content from documents. More specifically, embodiments of the technology relate to using fuzzy regular expressions to process content results obtained from one or more documents in order to extract content for these documents. Further, embodiments of the technology enable the format modification of the content after the content has been identified and extracted from the documents.