Palm Image Affine Normalization for Biometric Authentication

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In biometric authentication, especially palm vein authentication, position differences between input and enrollment biometric information can lead to authentication errors, particularly for users with larger hands, where parts of the hand may not be captured, increasing error rates and usability issues, especially with mobile devices.

Innovation Solution

An image processing device captures a palm image, detects base points of specific fingers, calculates standard points, and applies affine conversion to normalize the image, ensuring consistent coordinates for stable authentication, even if parts of the palm are not included in the image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a guide for guiding hand posture is used to reduce position difference influence, then authentication accuracy is improved, but users with larger hands cannot have their entire hand captured, increasing authentication error rate

Engineering Contradiction:

Improveauthentication accuracyVSAvoidauthentication error rate

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The patent changes the parameter of coordinate system transformation by detecting base points of fingers and applying affine conversion to transform palm images to standardized coordinates, accommodating users with larger hands while maintaining authentication accuracy

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent introduces base point detection of fingers as an intermediary element to establish reference points for coordinate transformation, enabling accurate positioning without requiring the entire hand to be captured

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If affine conversion is applied to normalize palm image coordinates, then position gap between input and enrollment biometric information is reduced, but image processing complexity increases

Engineering Contradiction:

Improveposition alignment accuracyVSAvoidimage processing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary detection of base points of fingers before authentication comparison, establishing reference coordinates in advance to simplify the affine conversion process and reduce real-time processing complexity

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS10438078B2Image processing device, image processing method and computer-readable non-transitory medium

Publication Date: 2019.10.08 FUJITSU LTD

AI summary

An image processing device includes: a biometric sensor; and a processor to execute a process comprising: detecting base points of a first and a second fingers; calculating a first standard point between one of the base points of the first finger and one of the base points of the second finger; calculating a second standard point on a bisector of an angle formed by a straight line passing through one of the base points of the first finger on an opposite side of the second finger and the first standard point and a straight line passing through one of the base points of the second finger on an opposite side of the first finger, and the first standard point; and applying an affine conversion to the palm image so that coordinates of the first and second standard points are converted into predetermined coordinates.