Image Registration via Normalized Coordinate Intermediary

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image processing technologies face challenges in accurately registering deformed medical images, as coordinate transformation can distort the object shape, making it difficult to identify corresponding points and perform precise registration.

Innovation Solution

An image processing apparatus that includes a coordinate transformation unit, an information obtaining unit, and a deformation transformation unit to transform images into a normalized space for easy comparison, allowing users to identify corresponding points and perform registration based on original image features.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If coordinate transformation is applied to transform images to a normalized space, then it is facilitated to identify corresponding points, but the shape of the object in the transformed image becomes significantly different from the original image, making it difficult to perform registration

Engineering Contradiction:

Improveease of identifying corresponding pointsVSAvoidregistration accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent uses a normalized coordinate space as an intermediary to facilitate corresponding point identification. Corresponding points are first identified in the normalized space where shapes are standardized, then transformed back to original image coordinates to perform registration. This intermediary approach allows easy point identification while maintaining registration accuracy in the original image space.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent introduces a normalized coordinate system as an additional dimensional framework. By mapping images to this standardized coordinate space, corresponding points become easier to identify visually. The solution then transforms these points back to the original image coordinate systems, effectively using the normalized space as a temporary dimensional bridge for accurate registration.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Difficulty of detecting and measuring

If coordinate transformation is applied to each image to facilitate identification of corresponding points, then visual observation is improved, but the deformed images become difficult to register with each other

Engineering Contradiction:

Improvedifficulty of identifying corresponding pointsVSAvoidregistration reliability

Core Design Contradiction:

Difficulty of detecting and measuringVSReliability

Solution Approach 1:

The normalized coordinate space serves as a mediator that temporarily standardizes image shapes for easy corresponding point detection. Once points are identified in this intermediary space, the system transforms them back to the original image coordinates, ensuring that registration is performed on the unreduced, reliable original images rather than on the deformed normalized versions.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent performs preliminary coordinate normalization to facilitate corresponding point identification, then transforms the identified points back to original coordinates before performing registration. This preliminary action of normalization is used only for point detection, not for the actual registration process, thereby maintaining registration reliability.

Inventive Principle:

Principle #10Preliminary action

3Ease of operation

If images are deformed to facilitate corresponding point identification, then visual comparison is improved, but the original image features are distorted, reducing registration accuracy

Engineering Contradiction:

Improvevisual comparison easeVSAvoidregistration precision

Core Design Contradiction:

Ease of operationVSManufacturing precision

Solution Approach 1:

The patent uses normalized coordinate space as a visual intermediary that preserves the ability to perform accurate registration. Corresponding points are identified in the normalized space where visual comparison is easy, but the actual registration transformation is calculated using the original unnormalized images, thus preserving manufacturing precision while maintaining operational ease.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

Instead of deforming the original images for registration, the patent inverts the approach by normalizing coordinates only for corresponding point identification. The registration process then proceeds in reverse, using the original image geometries to ensure precision, while benefiting from the ease of point identification provided by the normalized coordinate system.

Inventive Principle:

Principle #13The other way round (Inversion)

Data Source

PatentUS10762648B2Image processing apparatus, image processing method, image processing system, and program

Publication Date: 2020.09.01 CANON KK

AI summary

An image processing apparatus in which identification of a corresponding target point between images is easily performed, and physically appropriate registration can be performed by using the identification is provided. The image processing apparatus obtains coordinate transformation for respectively transforming a first image and a second image obtained by imaging an object to a first transformed image and a second transformed image, obtains correspondence information that serves as information of a corresponding point included in each of the first transformed image and the second transformed image, respectively transforms the correspondence information to positions in the first image and the second image, and obtains deformation transformation for performing registration of the first image and the second image on the basis of information of the corresponding points in the first image and the second image.