Fingerprint Representation via Multi-Setting Feature Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Capacitive fingerprint sensors face challenges with humidity and pressure variations, leading to image saturation and reduced image quality, which complicates fingerprint verification and requires substantial computational power for image combination.

Innovation Solution

A method that acquires fingerprint images at different acquisition settings, extracts feature sets, and combines them to form a candidate fingerprint representation, reducing computational power and improving image quality by selecting features based on specific criteria, such as image salience and contrast.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple fingerprint images are captured at different sensitivity levels and combined, then image quality and verification reliability are improved, but computational power requirements and processing time increase substantially

Engineering Contradiction:

Improvefingerprint verification reliabilityVSAvoidcomputational power requirement

Core Design Contradiction:

ReliabilityVSPower

Solution Approach 1:

The patent extracts only the essential fingerprint features from multiple images at different sensitivity levels, rather than processing and combining the complete images. This selective extraction of relevant feature data reduces the computational burden while maintaining verification reliability.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the fingerprint verification process into distinct stages: capturing multiple images at different sensitivity levels, extracting features from each image, and combining only the extracted feature sets. This segmentation allows computational resources to be focused on processing compact feature representations rather than large image data.

Inventive Principle:

Principle #1Segmentation

2Reliability

If multiple fingerprint images are captured at different sensitivity levels and combined, then image quality and verification reliability are improved, but processing time increases

Engineering Contradiction:

Improvefingerprint verification reliabilityVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs feature extraction as a preliminary action before combining multiple images. By extracting features first and then combining only the compact feature sets rather than full images, the processing time is significantly reduced while still achieving improved verification reliability through multi-level sensitivity data.

Inventive Principle:

Principle #10Preliminary action

3Power

If feature representation sets are combined in feature space rather than image space, then computational power requirements are reduced, but implementation complexity increases

Engineering Contradiction:

Improvecomputational power requirementVSAvoidfeature processing complexity

Core Design Contradiction:

PowerVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical approach of combining full images with a more efficient mathematical approach of combining feature representation sets in feature space. This substitution uses abstract mathematical representations (feature vectors) instead of concrete image pixel data, reducing computational requirements while the structured feature extraction process manages the implementation complexity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS10037454B2Method and device for forming a fingerprint representation

Publication Date: 2018.07.31 FINGERPRINT CARDS IP AB

AI summary

The present invention generally relates to a method for forming a candidate fingerprint representation of a fingerprint of a finger of a user of an electronic device comprising a capacitive fingerprint sensor for sensing a fingerprint pattern, the method comprising acquiring a sequence of candidate fingerprint images each acquired at different imaging acquisition settings. Further, a fingerprint feature set is extracted from each of the candidate fingerprint images, the fingerprint features of each fingerprint set being extracted according to a first selection criterion. A feature representation set is determined for each of the fingerprint feature sets. A candidate fingerprint representation is formed from a combination of the fingerprint feature representation sets. The invention further relates to a fingerprint sensing system and an electronic device.