| Session P2：網路犯罪與數位鑑識
論文作者：Wen-Hui Lin
Every station of a public transportation system has numerous turnstiles employing radio frequency identification (RFID) and closed-circuit television (CCTV) cameras to collect customer identities and biometric information for managerial needs. The safeguarding of personal information privacy is an increasing concern in contemporary society. Accordingly, this study proposed a secret-splitting-based face recognition model (SFRM). In the scheme, the wavelet transforms are used to distribute a secret in two parts. Specifically, the proposed authentication scheme uses a part of the information authentication constructed from the approximation of the faceimage information stored in the RFID tag, whereas the other part of the detailed facial-feature information, which corresponds to the authorized approximation image, is extracted by local binary patterns (LBPs) and stored in a signature database for matching. The accuracy of user authentication and face recognition can be determined both by the nearest–neighbor (NN) approach and by two threshold values. Experimental results showed that our system provides a more secure technique for public-transportation security mechanisms and avoid disclosure of personal information. |
