Document Area Pattern Selection for Scanner Edge Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing scanner apparatuses with multi-crop functions face challenges in accurately extracting document areas, leading to errors in cutting multiple documents as one large document due to unclear document edges when black background paper is missed, and the complexity of user operations to correct erroneous document area extraction.

Innovation Solution

An information processing apparatus that displays multiple patterns of document areas extracted from an input image, allowing users to select the correct pattern, thereby simplifying the process of extracting accurate document areas and reducing user burden.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If black background paper is used to increase contrast, then document edge clarity is improved, but operation complexity increases and errors may still occur

Engineering Contradiction:
Improvedocument edge clarityVSAvoidoperation complexity
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system automatically extracts multiple document area patterns and presents them to the user for selection, eliminating the need for manual document repositioning or black background paper usage. The apparatus self-corrects by providing multiple extraction patterns including correct and erroneous ones, allowing users to simply select the correct pattern rather than performing complex manual operations.

Inventive Principle:
Principle #25Self-service

2Productivity

If multiple documents are placed close together or aligned, then productivity is improved, but document area extraction accuracy deteriorates

Engineering Contradiction:
Improvedocument processing efficiencyVSAvoiddocument area extraction accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system segments the document area extraction into multiple distinct patterns, presenting each possible extraction result as a separate option to the user. This segmentation allows the system to handle closely placed or aligned documents by providing multiple extraction patterns, enabling users to select the correct document boundaries even when documents are positioned close together or overlap slightly.

Inventive Principle:
Principle #1Segmentation

3Ease of operation

If simple display of extracted document area is used, then ease of operation is improved, but accuracy of document area extraction deteriorates

Engineering Contradiction:
Improveuser operation simplicityVSAvoiddocument area extraction accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The system provides feedback by displaying multiple extracted document area patterns to the user, allowing visual confirmation and selection of the correct extraction. This feedback mechanism maintains ease of operation while improving accuracy, as users can simply select from displayed patterns rather than performing complex manual corrections, ensuring only accurate document areas are processed further.

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS11196887B2Information processing apparatus receiving user selection among patterns to correct document areas and non-transitory computer readable medium storing program
Publication Date: 2021.12.07 FUJIFILM BUSINESS INNOVATION CORP
  • US11196887B2 patent drawing
  • US11196887B2 patent drawing
  • US11196887B2 patent drawing

AI summary

An information processing apparatus includes a processor configured to, in a case where there are plural patterns of document areas extracted from an input image obtained by imaging documents, execute display control for displaying the plural patterns as options on a screen, receive selection of a correct pattern among the plural patterns according to the display of the screen through the display control, and extract an image of each document area included in the pattern selected by a user from the input image.