OCR Page Division Using History for Mixed Scanned Documents
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image forming apparatuses struggle with accurately dividing and processing multiple scanned page data, particularly in identifying and organizing different types of form documents, as they often require manual rule creation and fail to utilize historical data for efficient page classification.
Innovation Solution
An information processing apparatus equipped with an OCR unit for character and layout recognition, a division history unit for storing page data classifications, and a rule order unit that utilizes historical data to automatically classify and divide page data into units based on layout and character analysis, assisted by machine learning when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual rule creation is used for dividing scanned page data, then flexibility in handling different document types is improved, but processing time and operational complexity increase
Solution Approach 1:
The system performs preliminary OCR recognition and classification on scanned page data to pre-categorize documents before division. By analyzing character patterns, layout structures, and document types in advance, the system prepares classification results that guide subsequent automatic division, eliminating the need for manual rule creation during actual processing and significantly reducing operational time
Solution Approach 2:
The division execution unit automatically divides scanned page data based on classification results from the OCR recognition unit, without requiring manual intervention. The system serves itself by using its own recognition and classification capabilities to drive the division process, adapting to different document types through automated pattern recognition rather than manual rule configuration
2Measurement precision
If manual rule creation is used for dividing scanned page data, then accuracy in document classification can be maintained, but ease of operation deteriorates
Solution Approach 1:
The system automatically performs OCR recognition, classification, and division of scanned page data without requiring manual rule creation or user intervention. The division execution unit autonomously processes classified documents based on recognition results, maintaining high classification accuracy through sophisticated pattern recognition while dramatically improving ease of operation by eliminating manual steps
Solution Approach 2:
The system uses classification results from OCR recognition as feedback to guide the division process. By continuously analyzing document patterns, layout structures, and content characteristics, the system adjusts its division strategy based on recognition accuracy, maintaining high precision while operating automatically without manual input
3Productivity
If historical data is not utilized, then system complexity remains low, but productivity in processing multiple documents deteriorates
Solution Approach 1:
The system performs preliminary OCR recognition and classification on scanned page data to pre-categorize documents before division. By analyzing character patterns, layout structures, and document types in advance, the system prepares classification results that guide subsequent automatic division, eliminating the need for manual rule creation during actual processing and significantly reducing operational time
Solution Approach 2:
The division execution unit automatically divides scanned page data based on classification results from the OCR recognition unit, without requiring manual intervention. The system serves itself by using its own recognition and classification capabilities to drive the division process, adapting to different document types through automated pattern recognition rather than manual rule configuration
Data Source
AI summary
Provided is an information processing apparatus that divides a plurality of scanned page data with high accuracy based on the history. The OCR unit performs optical character recognition for character and layout in a page for each of the plurality of page data. The division history unit stores the division history of a plurality of page data on a page-by-page basis based on the classification of characters and layouts that are performed optical character recognition by the OCR unit. The rule order unit classifies each of a plurality of newly scanned page data, and it divides the data in page units by referring to the division history stored in the division history unit.