OCR Verification System Using Formula-Based Data Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical character recognition (OCR) systems face challenges in accurately interpreting handwritten, degraded, or numerical documents, leading to potential errors that require significant human verification, especially in financial documents like tax forms, which can result in misinterpretation and require substantial manpower for validation.
Innovation Solution
A system and method for verifying and correcting OCR data using a processor and memory, where data fields are marked as uncertain or verified based on predefined formulas and rules, allowing automatic verification and correction without the need for prior documentation, minimizing human intervention and ensuring high accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If OCR processes are used to convert scanned paper documents into machine-readable format, then productivity and ease of document management are improved, but reliability and measurement precision deteriorate due to errors in deciphering handwriting, fonts, or degraded documents
Solution Approach 1:
The system applies feedback by using verified data fields to validate unverified fields through predefined formulas. The verification process continuously checks data relationships and provides corrective feedback when discrepancies are detected, improving overall OCR reliability while maintaining automated processing speed.
Solution Approach 2:
The patent introduces an intermediary verification layer between OCR processing and final data usage. This intermediary system uses predefined formulas and validation rules to check the accuracy of OCR output without requiring manual review of every field, thus maintaining productivity while improving reliability.
2Reliability
If multiple passes of different OCR technologies or human operators are used to verify OCR accuracy, then reliability improves, but productivity and device complexity worsen due to substantial time requirements
Solution Approach 1:
The verification system performs self-service by automatically validating OCR output against predefined formulas and business rules. The system checks data consistency and identifies errors without requiring external human operators or multiple OCR passes, thus improving reliability while maintaining high productivity.
Solution Approach 2:
Instead of verifying every single OCR field, the system applies partial verification by focusing on critical data fields and relationships defined by predefined formulas. This selective approach provides sufficient verification accuracy without the time cost of complete verification of all fields.
3Reliability
If human operators are employed to crosscheck and verify OCR accuracy, then reliability improves, but productivity and loss of time worsen due to significant human manpower requirements
Solution Approach 1:
The patent replaces the mechanical system of human operators with an automated verification system based on predefined formulas and validation rules. This substitution eliminates manual verification time while maintaining high reliability through systematic automated checking of data relationships.
Solution Approach 2:
The verification system performs self-service by automatically detecting and flagging errors in OCR output. The system validates data fields against business rules and formulas without requiring external human intervention, thus eliminating time loss associated with manual verification while maintaining high accuracy.
4Reliability
If prior OCR'd documents are used to establish confidence levels in subsequent OCR accuracy, then reliability improves, but device complexity and loss of time worsen due to requirements for substantial prior documentation
Solution Approach 1:
The verification system performs self-service by using predefined formulas and validation rules that are independent of prior OCR documents. The system validates current OCR output against established business logic without requiring comparison to previous documentation, thus maintaining reliability while reducing system complexity.
Solution Approach 2:
The system performs preliminary action by pre-defining validation formulas and business rules before OCR processing. These predefined criteria enable immediate verification of OCR output without requiring posterior comparison to prior documents, thus establishing confidence levels through advance preparation rather than historical comparison.
Data Source
AI summary
A system for verifying and correcting errors after translation of printed text into machine-readable text. The system includes a memory for storing formulas defining relationships between data fields. A processor evaluates the formulas according to data values associated with the data fields to determine whether the formulas evaluate as truthful statements. The processor marks the data fields of the formulas as unverified or as verified based upon this evaluation. The system also uses the processor to calculate a determined value for data fields in an attempt to correct errors in the translation of the printed text into machine-readable text. If different determined values are calculated for the same data field, based upon different formulas, the data field is marked as uncertain. The system iterates based upon the marking of the data fields of the formulas as verified or unverified and as uncertain or not uncertain.