Automated Retrieval File Generation for Document Image Processing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current document retrieval methods are inefficient as they rely on manual processing and lack effective automation for generating retrieval information, especially when dealing with large document files containing images and text, which hinders quick and accurate keyword extraction and visual summarization.

Innovation Solution

A retrieval information generation device that includes a determination unit for identifying document information, an extraction unit for extracting visual image information, and a generation unit that associates this information to create a retrieval file, enabling automated generation of retrieval files with visual recognition images and keyword information linked to the original document.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual processing methods are used for document retrieval, then operators can carefully examine each document, but the retrieval process becomes time-consuming and inefficient

Engineering Contradiction:

Improveretrieval accuracyVSAvoidretrieval time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the document processing task into distinct functional units: an extraction unit that separates image information from document files, a determination unit that identifies candidate phrases, and a verification unit that confirms keyword accuracy. This segmentation enables automated parallel processing while maintaining retrieval precision, resolving the contradiction between manual accuracy and automated speed.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary automated processing system that acts as a bridge between manual document examination and final retrieval decisions. The extraction unit, determination unit, and verification unit work together as intermediaries to pre-process and filter documents, reducing the time burden while preserving accuracy through multiple verification stages.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If automated extraction methods are implemented, then processing speed increases, but the precision of keyword extraction and visual summarization deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidkeyword extraction accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent implements a feedback mechanism where the determination unit generates candidate keywords based on extracted image information, which are then verified by a verification unit against the original document content. This feedback loop ensures that automated extraction maintains high precision by continuously validating results, resolving the contradiction between speed and accuracy.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The extraction unit performs preliminary action by pre-extracting image information and generating candidate keywords before the final verification step. This preliminary processing accelerates the overall workflow while the subsequent determination and verification steps ensure precision is maintained, allowing automated processing to achieve both speed and accuracy.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If comprehensive document analysis is performed, then retrieval quality improves, but the complexity of the processing system increases

Engineering Contradiction:

Improveretrieval qualityVSAvoidprocessing system complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent divides the comprehensive document analysis into modular segments: image extraction, candidate determination, and verification. Each module performs a specific function with well-defined inputs and outputs, making the overall complex system manageable and maintainable while achieving high retrieval quality through the coordinated operation of these segmented components.

Inventive Principle:

Principle #1Segmentation

4Adaptability or versatility

If visual image information is extracted and associated with document information, then visual recognition capability improves, but the time required for information aggregation increases

Engineering Contradiction:

Improvevisual recognition capabilityVSAvoidinformation aggregation time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The extraction unit performs preliminary action by extracting and preprocessing visual image information from document files before the association step. This preliminary extraction organizes the visual data in advance, enabling faster aggregation and association with document information later, thus improving visual recognition capability without proportionally increasing aggregation time.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS10445375B2Retrieval information generation device, image processing device, and non-transitory computer readable medium

Publication Date: 2019.10.15 FUJIFILM BUSINESS INNOVATION CORP

AI summary

A retrieval information generation device includes a determination unit that determines retrieval document information for retrieving a phrase in an original document from an original document file which is created by reading the original document by an image reading device and is identified on the basis of identification information, an extraction unit that extracts retrieval image information for visually specifying the original document from the original document file, and a generation unit that associates the retrieval document information and the retrieval image information with the identification information of the original document file to generate a retrieval file.