Camera Parameter Setting via Selectable 2D–3D Edge-Line Alignment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing camera parameter estimation methods struggle with accuracy due to the complexity of the physical environment, making it difficult to align 3D model ridge lines with object edges accurately, especially when adjusting multiple parameters manually.

Innovation Solution

An information processing device that assists users in calibrating camera parameters by displaying selectable edge lines from both the captured image and 3D model, allowing manual adjustment through guided procedures to align feature lines with projected ridge lines, using a combination of rotation, focal length, and translational movements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If camera parameters are estimated automatically from environmental objects, then the process is automated, but the accuracy deteriorates due to environmental complexity

Engineering Contradiction:
Improvecamera parameter estimation automationVSAvoidcamera parameter accuracy
Core Design Contradiction:
Extent of automationVSMeasurement precision

Solution Approach 1:

The patent introduces a reference object with known geometric features as an intermediary between the camera and environmental objects. This reference object serves as a mediator that provides stable, controlled feature lines for accurate camera parameter estimation, eliminating the need to rely on complex environmental objects while maintaining automation.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent uses a virtual reference object (computer-generated model) that replicates the physical reference object's geometric features. This virtual copy provides idealized, noise-free feature lines that can be precisely matched with captured image features, ensuring high accuracy in camera parameter estimation without being affected by environmental complexities.

Inventive Principle:
Principle #26Copying

2Measurement precision

If multiple camera parameters are adjusted manually to align 3D model ridge lines with object edges, then the alignment accuracy improves, but the operation complexity increases

Engineering Contradiction:
Improvealignment accuracyVSAvoidparameter adjustment difficulty
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent implements a feedback mechanism where the system automatically calculates alignment errors between 3D model ridge lines and captured image edge lines, then guides the user to adjust camera parameters to reduce these errors. This feedback loop enables accurate alignment while simplifying user operation by providing clear adjustment guidance.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent replaces manual trial-and-adjustment mechanical operations with automated computational geometry processing. The system uses algorithmic calculations to determine optimal camera parameters based on feature line matching, substituting complex manual mechanical adjustment with automated mathematical computation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Adaptability or versatility

If feature lines are extracted from environmental objects, then the system adapts to real-world scenes, but the measurement difficulty increases due to object complexity

Engineering Contradiction:
Improvescene adaptation capabilityVSAvoidfeature line detection difficulty
Core Design Contradiction:
Adaptability or versatilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent introduces a controlled reference object as an intermediary that simplifies feature line detection. Instead of extracting features from complex environmental objects, the system uses the reference object's known geometric features as a mediator, making detection easier while still adapting to real-world scenes through the reference object placement.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS20250291976A1Non-transitory computer-readable recording medium, setting method, and information processing device
Publication Date: 2025.09.18 FUJITSU LTD
  • US20250291976A1 patent drawing
  • US20250291976A1 patent drawing
  • US20250291976A1 patent drawing

AI summary

An information processing device includes a processor, the processor is configured to execute processing of acquiring a two-dimensional video in which an object is arranged in a physical space, the two-dimensional video being captured by a camera device, acquiring three-dimensional design data defining formation of the object in a virtual space, displaying, on a screen, in a selectable state, first edge lines of the object extracted from a captured image included in the two-dimensional video and second edge lines of the object in the three-dimensional design data, receiving a selection instruction indicating a pair of the first edge line and the second edge line that are targets to be superimposed, and setting camera parameters of the camera device based on superimposition operation for the pair of the first edge line and the second edge line.