Biometrics also needs safety technology

So far, any biometric system or method has the potential to make mistakes. To give an objective and accurate assessment of the system's recognition accuracy is actually a very complex issue. It is affected by the number of test samples, quality, evaluation indicators and other factors, but it is a concern for the application unit and the judicial department. Focus. Therefore, the performance evaluation of biometric identification methods has become an important direction for biometric identification research. For the 1:1 identity verification system, there are two kinds of errors: one is to identify different people's biometrics as the same type, which is called error reception; the other is to identify the same person's biometrics as different types. , called false rejection.

Generally, the performance index of a biometric method can be evaluated from both theoretical and experimental aspects. From the aspect of theory, we can study the uniqueness of biometrics, that is, accurately model the various parameters that affect wrong reception and false rejection, and give the theoretical error rate that can be obtained from the nature and mechanism of each biometric identification method. Lower bound. This work is very meaningful and difficult. For example, the judiciary has a lot of controversy on identifying criminals through fingerprint matching results. Although researchers have claimed that two people with identical fingerprint characteristics cannot be found on the earth, how large does it need to be in an automatic or manual fingerprint recognition system? Can similarity completely confirm the homology of two fingerprints? What exactly is the exact probability of error recognition? Researchers have conducted more in-depth research on this issue, but this issue has not been completely resolved.

Biometrics also needs security technology. A biometric system is a security system. Like other information security technologies, it can also be subject to various attacks and can be threatened by hackers at every stage of the identification system. In addition to forging other people's biometric samples, other possible attacks include: modifying the sample data on the communication link between the acquisition device and the computer, modifying the recognition result, replacing the matching program, and attacking the biometric template database. And each link's attack is fatal, so the safety factor of the biometric system depends on the weakest link.

In order to make biometrics technology available for applications with high security requirements, in addition to algorithm design, it is also important to protect the system's own security and improve its resistance to various hacking attacks. In order to improve the security of the recognition system, it is a feasible research direction to adopt methods such as encryption, digital signature, and time stamping for biometric data, feature templates and applications.