Fingerprint Authentication Using Variation Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fingerprint authentication methods in smartphones require users to place their fingers in predetermined positions and stretch them out, which decreases user comfort and slows down the screen unlocking process, as they rely on isolating individual fingerprints and verifying their geometry.
Innovation Solution
A method that uses a fingerprint sensor to acquire and process multiple fingerprints simultaneously, employing variation models to determine plausible variations in finger positions and update them based on similarity and plausibility, allowing for more flexible and efficient authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fingerprint data of different fingers are acquired and individual fingerprints are isolated and geometry is verified, then authentication reliability is improved, but user comfort deteriorates and authentication speed decreases
Solution Approach 1:
The patent segments the fingerprint authentication process by treating each finger independently through individual isolation and geometry verification, while simultaneously processing multiple fingerprints in parallel. This segmentation allows reliable authentication through detailed feature analysis while improving user comfort by enabling natural, uncoordinated finger placement on the sensor surface.
2Reliability
If fingerprint data of different fingers are acquired and individual fingerprints are isolated and geometry is verified, then authentication reliability is improved, but authentication speed decreases
Solution Approach 1:
The patent maintains continuous useful action by processing multiple fingerprint images in parallel without requiring sequential isolation or geometry verification of each finger. The system continuously captures and processes fingerprint data from multiple fingers simultaneously, maintaining high authentication speed while preserving reliability through comprehensive feature analysis of all detected fingerprints.
3Measurement precision
If users are required to place fingers in predetermined positions and stretch them out, then measurement precision is improved, but ease of operation deteriorates
Solution Approach 1:
The patent applies dynamics by making the fingerprint sensor adaptable to various finger configurations rather than requiring fixed positions. The sensor dynamically processes fingerprints from fingers in any position or configuration on the sensor surface, maintaining measurement precision through flexible image processing algorithms that can isolate and analyze individual fingerprints regardless of their initial placement.
4Ease of operation
If multiple fingerprints are processed simultaneously without isolation, then user comfort is improved, but measurement precision deteriorates
Solution Approach 1:
The patent segments the processing of multiple fingerprints by automatically isolating individual fingerprint images from the captured data, even when fingers are in natural, uncoordinated positions. This segmentation enables precise feature extraction and geometry verification of each finger while maintaining user comfort through flexible, position-independent acquisition.
Data Source
AI summary
A method to authenticate a user by means of a fingerprint sensor, which is capable of simultaneously acquiring the fingerprints of a hand, the method comprising: an enrollment process (100), during which the features of first fingerprints (Ei) of the fingers of the hand are acquired (101-103), a variation model (Gij) indicating a plausible mutual position variation for each pair of first fingerprints (Ei, Ej) is determined (104) based on the respective features and stored (105); and a comparison process (200), during which features of successive fingerprints (Ei, Ej) of the fingers of the hand are acquired (201-203), a degree of similarity between each one of the first fingerprints (Ei) and each one of the successive fingerprints (Ck) is determined (204) based on a comparison of the respective features, a degree of plausibility of mutual position of pairs of fingerprints (<Ei,Cki>, <Ej,Ckj>) of a set (P) of pairs of fingerprints, each comprising one of the first fingerprints (Ei) and that of the successive fingerprint (Cki) offering the highest degree of similarity, is determined (207, 208) based on the features of the pairs of fingerprints and on the variation model (Gij) of the first fingerprints (Ei, Ej) of the pairs of fingerprints, the user is recognized (209-211) and the variation models (Gij) are updated (212) based on the degrees of similarity and on the mutual position plausibility degrees.