Font Size: a A A

Secure Fingertip Mouse Authentication System Based On Fingertip Tracking For Mobile Devices

Posted on:2017-11-30Degree:MasterType:Thesis
Country:ChinaCandidate:Q ChenFull Text:PDF
GTID:2348330491962674Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
In recent years, touch-screen smart devices rapidly permeate almost every aspect of people's life, including social activity, consumption, Internet surfing and so on. A lot of users' sensitive information bundles together with smart devices. Therefore a variety of attackers try to crack users' authentication process on smart devices in order to access the user's sensitive information. Current various attacks on smart device may disclose sensitive information such as passwords of mobile devices. These attacks primarily consist of three types of side channel attacks, including computer vision-based attacks, residue-based attacks, and sensor based attacks. Computer vision based attacks analyze the interactions between fingers and the touch screen, residue-based attacks exploit oily or heat residues on the touch screen, and sensor-based attacks silently monitor and analyze the data from diverse motion sensors so as to infer the tapped keys.In this paper, we introduce a novel authentication system, referred to as "Secure Finger Mouse", to defeat these side channel attacks. A user can input her/his password by moving her/his finger and clicking in front of the rear camera of the smart device without physically touching the screen. Our main work includes:(1) for smart device with single camera, the system use single camera to record and analyze user's fingertip motion in order to identify the fingertip movement and click gesture by computing its 2D information through image processing. By tracking fingertip actions of users, the user could control the movement of an on-screen cursor and perform key selection and input using a soft keyboard. (2) For smart device with dual cameras, the system records user's fingertip motion and computes 3D information to identify its movement and click gesture. The same principles are applied to control cursor on the screen and enter passwords. (3) The prototype system of secure finger mouse is implementted on the Android smartphone and a stereo camera simulation system, respectively. Then we analyze and experimentally evaluate usability and security of the system in order to demonstrate the feasibility and effectiveness of our system. The system could effecitively defeat residue-based attacks and sensor-based attacks since users don't touch the screen. Moreover, our system introduces random parameters in mouse movement so as to break the mapping relationship between the physical space movement and display space movement. Thus, it could defeat computer vision based attacks and sensor-based attacks as well.In summary, to defeat computer vision based attacks, residue-based attacks, sensor-based attacks, we design a fingertip tracking based authentication system referred to as "secure finger mouse". The system is implemented on the Android smartphone and a stereo camera simulation system respectively. It can provide security protection for user authentication.
Keywords/Search Tags:Side channel attacks, Image processing, Fingertip detection, Stereo vision, Click estimate
PDF Full Text Request
Related items