Font Size: a A A

Research On Multi-modal Finger Recognition Technology

Posted on:2021-03-01Degree:MasterType:Thesis
Country:ChinaCandidate:Z G TaoFull Text:PDF
GTID:2428330632950623Subject:Optical Engineering
Abstract/Summary:PDF Full Text Request
Nowadays,personal authentication is becoming more and more common in many scenarios.Biometrics,which is represented by fingerprints and human faces,have been widely used in many fields for convenience and reliability.Emerging biometric technologies such as iris,vein,and voice also develop quickly.However,each biometrics has its own advantages and disadvantages.Therefore,developing multi-modal biometric system which usually combines different biometric features has become the consensus of the industry.The finger is the richest part in human-specified information of the body.The vein,fingerprint,and inner knuckle print all have high specificity and can be used as personal identity information.In this paper,the multi-modal biometric of finger veins and inner knuckle print was studied.The work in this paper mainly includes the following contents:1.A finger vein scanner system has been built.The system is user-friendly and has the features of small size,high imaging quality,adaptive adjustment of light intensity.It can collect finger vein images in real time.And we also have established a finger vein database based on the device for algorithmic research.2.We have researched on the image processing algorithm of finger vein in depth.And a center line matching method with local coding is proposed,which improves the accuracy and reliability of finger vein extraction,and the flexibility of feature matching,and improves the accuracy of finger vein recognition.3.An inner knuckle print scanner system has been built to improve the imaging quality of the inner knuckle print.Based on the device,an inner knuckle print image database has been established for the algorithm research.4.We have researched on the image processing algorithm of inner knuckle print in depth.A texture extraction algorithm based on Hessian matrix is proposed,which improves the accuracy of feature extraction.Based on the experimental database,the performance is analyzed and high accuracy is achieved,which verifying the validity of the inner knuckle print as a biometric information5.Based on the databases of finger vein and inner knuckle print,an experiment has been conducted on the fusion identification of finger vein and inner knuckle print.The result shows that the characteristics of multimodal fusion can significantly improve the accuracy of biometric system.
Keywords/Search Tags:Computer vision, Biometrics technology, Finger vein recognition, Inner-knuckle-print recognition, Multi-modal biometrics technology
PDF Full Text Request
Related items