Document Inclination Detection Using Shadow and Edge Boundaries

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing electric skew correction techniques are inaccurate due to the influence of document states like folding and curling, and document color or density, leading to erroneous detection of skew and registration amounts.

Innovation Solution

The method involves using multiple inclination detecting units to analyze the boundary between document shadows and edges, combining detection results to accurately determine the inclination amount, and applying correction only when necessary to minimize errors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If shadow detection is used for inclination measurement, then detection can be performed on documents with color or density differences, but accuracy deteriorates due to document states like folding and curling affecting shadow formation

Engineering Contradiction:
Improvedetection capabilityVSAvoidinclination detection accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments the inclination detection process into two independent parts: shadow boundary detection and document edge detection. By detecting the shadow boundary first to determine the detection range, then detecting the document edge within that range, the system separates the functions to leverage the advantages of both methods while mitigating their individual disadvantages.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces the shadow boundary as an intermediary element. The shadow boundary detection serves as a mediator that defines the detection range for subsequent document edge detection. This intermediary approach allows the system to use shadow information adaptively without being constrained by its inaccuracies, enabling precise document edge detection even on documents with folding or curling.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of operation

If document edge detection is performed without shadow boundary reference, then direct document outline detection is possible, but accuracy deteriorates due to documents with similar color or density to the background

Engineering Contradiction:
Improvedetection process simplicityVSAvoiddocument outline detection accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent performs preliminary shadow boundary detection before document edge detection. By first identifying the shadow boundary to establish the detection range, the system prepares the detection framework in advance. This preliminary action ensures that subsequent document edge detection is confined to the appropriate range, improving accuracy for documents with low contrast against the background.

Inventive Principle:
Principle #10Preliminary action

3Manufacturing precision

If electric skew correction is applied based on detected skew amount, then document alignment is improved, but erroneous correction occurs when detection accuracy is compromised by document states

Engineering Contradiction:
Improvedocument alignment precisionVSAvoidcorrection reliability
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent segments the correction reliability assessment into multiple independent detection results: shadow boundary detection result and document edge detection result. By evaluating these separate results independently and comparing them, the system can identify inconsistencies caused by document states like folding or curling, thereby maintaining high correction reliability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements a feedback mechanism where the shadow boundary detection result serves as a reference to evaluate the document edge detection result. By comparing these results and determining whether their difference is within a predetermined range, the system provides feedback on detection quality, enabling reliable decision-making about whether to apply skew correction.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3691237B1Inclination detecting device, reading device, image processing apparatus, and method of detecting inclination
Publication Date: 2025.08.27 RICOH CO LTD
  • EP3691237B1 patent drawingFigure 1
  • EP3691237B1 patent drawingFigure 2
  • EP3691237B1 patent drawingFigure 3

AI summary

An inclination detecting device includes a first boundary detecting unit (310), a second boundary detecting unit (320), and an inclination detecting unit (340). The first boundary detecting unit (310) is configured to detect, in image information that is an image of an object imaged by an imaging device (28) at an imaging position where a background member (92) forms a background, a first boundary between the background member (92) and a shadow of the object. The second boundary detecting unit (320) is configured to detect a second boundary between the object and the shadow of the object in the image information. The inclination detecting unit (340) is configured to detect an inclination of the object in the image information from the first boundary detected by the first boundary detecting unit and the second boundary detected by the second boundary detecting unit.