Show simple item record

dc.contributor.advisorAsaduzzaman, Abu
dc.contributor.authorMummidi, Abhishek
dc.date.accessioned2015-07-23T16:46:55Z
dc.date.available2015-07-23T16:46:55Z
dc.date.issued2014-12
dc.identifier.othert14076
dc.identifier.urihttp://hdl.handle.net/10057/11349
dc.descriptionThesis (M.S.)--Wichita State University, College of Engineering, Dept. of Electrical Engineering and Computer Science
dc.description.abstractStandard networking protocols can be very vulnerable to spoofing attacks when extra precautions are not taken by applications to verify the identity. There are several ways to spoof the facial recognition using photograph, video, and three-dimensional (3D) model of a valid user. Studies show that there have been significant improvements in detecting photograph, video, and 3D model spoofing individually, but no such improvement in detecting all types of spoofing in a single design. Recent studies also show that liveness monitoring using the facial features has potential to improve security, especially in biometric systems. In this work, an omproved method using acoustics and liveness monitoring system to better detect spoofing is introduced. The proposed system includes an ultrasonic range sensing module, a light dependent resistor (LDR) for detecting light intensity, and an eye blinking password is developed for liveness detection of the users. Eye blinking password is generated with the help of local facial features such as eye blinking and chin movement pattern using a real-time generic web-camera. The improper lighting conditions are corrected by placing an external light source to provide suitable lighting for the system. The proposed system is tested by conducting experiments using 20 valid users and 80 other users with variations in user's expressions. According to the experimental results, the proposed system achieves 100% face recognition accuracy and 100% liveness detection performance.
dc.format.extentxii, 52 p.
dc.language.isoen_US
dc.publisherWichita State University
dc.rightsCopyright 2014 Abhishek Mummidi
dc.subject.lcshElectronic Thesis
dc.titleA novel spoofing detection methodology using acoustics and liveness monitoring system
dc.typeThesis


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record