High Confidence Visual Recognition of Persons by a Test of Statistical Independence
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.00 |
Reliable personal identity authentication has become an important issue in a number of wireless contexts including mobile commerce, access control, and user interface (UI) control. Traditional authentication approaches such as PIN codes or password queries are facing serious security-related problems. Biometrics-based authentication techniques make use of either physiological or behavioural traits that are able to differentiate between an authorized person and a fraudulent impostor. In this paper, both biometric technology and wireless setting-based constraints that determine the feasibility and the performance of the authentication feature are specified. The most well-known biometric approaches are briefly reviewed and their feasibility for wireless use is appraised. Furthermore, a number of quantitative and qualitative parameters for evaluation are also presented. Biometric technologies are continuously advancing toward commercial implementation in wireless devices. When carefully designed and implemented, the advantage of biometric authentication arises mainly from increased convenience and coexistent improved security.