Dual-Mode Image Processing for OCR Accuracy and Visual Quality

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for converting electronic image documents to searchable text files often compromise between accuracy and visual appeal, resulting in inefficient storage and retrieval, and may fail to locate documents due to errors in the OCR process, especially when dealing with colored or visually optimized images.

Innovation Solution

A system that creates a text-searchable data structure by generating both detail and visually optimized electronic image documents from physical or electronic images, using an OCR engine to abstract character information and link it with metadata for efficient text-based searching, while presenting visually optimized images to users.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a detail optimized electronic image document is created to maximize OCR accuracy, then the accuracy of text conversion is improved, but the file size increases and visual appeal deteriorates

Engineering Contradiction:

ImproveOCR accuracyVSAvoidfile size

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent segments the electronic image document into two separate versions: a detail optimized version for OCR processing and a visually optimized version for user display. This segmentation allows each version to be tailored for its specific purpose, resolving the contradiction between OCR accuracy and visual appeal while managing file size efficiently.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by creating different optimization levels for different uses of the same document. The detail optimized version has enhanced edge definition and contrast specifically for text regions to improve OCR accuracy, while the visually optimized version maintains overall image quality and color fidelity for user viewing, thus resolving the contradiction between OCR accuracy and visual appeal.

Inventive Principle:

Principle #3Local quality

2Measurement precision

If a detail optimized electronic image document is created to maximize OCR accuracy, then the accuracy of text conversion is improved, but storage requirements increase

Engineering Contradiction:

ImproveOCR accuracyVSAvoidstorage requirements

Core Design Contradiction:

Measurement precisionVSLoss of energy

Solution Approach 1:

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies parameter changes by adjusting image parameters such as edge definition, contrast, and resolution specifically for the detail optimized version to enhance OCR accuracy. The visually optimized version maintains different parameters for user viewing, thus resolving the contradiction between OCR accuracy and storage requirements through selective parameter optimization.

Inventive Principle:

Principle #35Parameter changes

3Ease of operation

If text files are created from electronic image documents using OCR to enable searching, then text-based search capability is improved, but errors in character recognition may cause documents to be missed

Engineering Contradiction:

Improvesearch capabilityVSAvoidsearch accuracy

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The patent applies preliminary action by creating a detail optimized electronic image document before OCR processing to maximize the accuracy of text conversion. This preliminary optimization of the image quality ensures that the subsequent OCR process produces more accurate text files, thereby improving search reliability while maintaining search capability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback by using the detail optimized electronic image document to generate text files with higher accuracy. The improved OCR accuracy provides feedback that reduces errors in the text files, ensuring that search results more reliably reflect the actual content of the documents, thus resolving the contradiction between search capability and search accuracy.

Inventive Principle:

Principle #23Feedback

4Ease of operation

If visual optimization is applied to electronic image documents to maintain appeal, then user satisfaction is improved, but OCR accuracy deteriorates

Engineering Contradiction:

Improveuser satisfactionVSAvoidOCR accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS7773822B2Apparatus and methods for management of electronic images

Publication Date: 2010.08.10 COLORMAX

AI summary

Systems including hardware and computer software and methods can create a text-searchable data structure that includes electronic image documents. The system may be configured in modules. The system converts an electronic image document into a visually optimized electronic image document and into a detail optimized electronic image document. The system also includes an OCR engine that abstracts character information from the detail optimized electronic image document and writes the character information into a text file. The visually optimized electronic image document is linked with the text file in a data structure by the system. The resulting data structure, which may be an image over hidden text pdf document, may be searched using various text based search techniques. When specified text is located in a text file, the corresponding visually optimized electronic image document may then be presented to the searcher.