Font Size: a A A

The Research Of Biometrics Identification Based On Deep Learning In APP

Posted on:2018-11-28Degree:MasterType:Thesis
Country:ChinaCandidate:S J LiFull Text:PDF
GTID:2348330518463686Subject:Engineering
Abstract/Summary:PDF Full Text Request
The modern software systems need more and more efficient,reliable and secure identification system with the development of the technologies.The use of traditional passwords for identity identification can be easily falsified,misappropriated and decoded,and that has failed to meet the safety requirements of modern applications.The password is simple but easy to crack,and it is strong but difficult to remember when user login the system uses the password authentication.Compared to traditional passwords,biometrics technology has no need for memory,and not easily lost,especially easy to use.Nevertheless the face recognition technology for system certification easy by the light,expression and occlusion of the impacts;Verification of Voiceprints technology easy is influenced by the noise interference and recording theft;Fingerprint recognition technology also exists the danger of fingers mold misappropriation and finger mold cracking and other issues.So,biometric identification technology also has a security risk in vivo detection.Therefore,a single mode of biometrics technology exists in vivo detections and other security risk,therefore,to achieve a fusion of variety of biometric architecture has important practical significance.In view of the above problems,the specific work of this thesis is as follows:(1)The biometrics security authentication architecture is designed based on deep learning that using face,voiceprint and fingerprint;(2)A highly secure identity authentication solution is designed on the security authentication architecture that using face detection and recognition technology,voice detection and voiceprint verification technology and fingerprint recognition technology;(3)The intelligent user interaction scheme is designed that based on GUI,speech recognition and speech synthesis technology.The variety of biological characteristics of the integration of security certification program is applied on the shopping APP security payments.The development of the relevant APP project shows the scheme is reliability and practicality.In further study,the program may be applied with bicycle sharing and logistics management.
Keywords/Search Tags:Face Recognition, Verification of Voiceprints, Fingerprint Recognition, Deep Learning, APP
PDF Full Text Request
Related items