Bounding Box Merging via Context-Aware Distance Thresholds

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing optical character recognition (OCR) systems generate bounding boxes that often require manual processing or customized software to meet specific organizational needs, such as word-level or paragraph-level bounding, which is costly and time-consuming.

Innovation Solution

A system that uses a processor and computer-readable medium to access documents, determine text characteristics, set distance thresholds based on context information using machine learning, and merge adjacent bounding boxes that satisfy these thresholds, adjusting thresholds dynamically based on context identifiers and accuracy information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If generic OCR software is used, then implementation cost is reduced, but bounding box output does not meet specific organizational needs

Engineering Contradiction:

Improveimplementation costVSAvoidbounding box output adaptability

Core Design Contradiction:

Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The system dynamically adjusts the distance threshold parameter based on document context information (such as document type, language, and layout characteristics) to optimize bounding box merging for different organizational needs without requiring custom software development

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the distance threshold parameter adaptively based on context information to achieve different bounding box merging outcomes (word-level, line-level, or paragraph-level) using the same generic OCR software

Inventive Principle:

Principle #35Parameter changes

2Manufacturing precision

If manual processing of bounding boxes is performed, then bounding box output meets specific needs, but time consumption increases

Engineering Contradiction:

Improvebounding box precisionVSAvoidprocessing time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The system performs automatic bounding box merging by evaluating distance thresholds and context information, eliminating the need for manual processing while achieving precise bounding box output that meets specific organizational needs

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system uses context information from the document (such as document type and layout characteristics) to automatically adjust merging parameters and achieve precise bounding box output without manual intervention

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If customized software is employed, then bounding box output meets local needs, but implementation cost increases

Engineering Contradiction:

Improvebounding box output adaptabilityVSAvoidimplementation cost

Core Design Contradiction:

Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

The system provides universal bounding box merging functionality that adapts to different organizational needs (word-level, line-level, paragraph-level) through dynamic parameter adjustment based on context information, eliminating the need for separate customized software solutions

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system dynamically adjusts the distance threshold parameter based on document context information to optimize bounding box merging for different organizational needs using the same generic OCR software

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS11670102B2Efficient bounding box merging

Publication Date: 2023.06.06 PAYPAL INC

AI summary

A system can merge text bounding boxes such as Optical Character Recognition (OCR) bounding boxes. A document can comprise a plurality of the text bounding boxes. Distance thresholds between text bounding boxes can be utilized for comparison against a distance threshold. Distance thresholds can vary depending on context information associated with the document. In response to a determination that text bounding boxes satisfy the distance threshold, the text bounding boxes can be assigned to a bounding box group.