| With the continuous progress of science and technology, an increasing number of electronic devices continue to enter the people's daily lives. Safe and user-friendly identity authentication technology have become increasingly demanding. If it has not very good identification and authorization system will not be able to carry out a very good e-commerce. There is too much reliance on more and more like a smart card, ID card, passwords and other protective measures to carry out identification of people, but all kinds of losses have occurred from time to time. Many of the existing computer systems and many have adopted the system of confidential user identification (ID) method to add a password to the user's identity authentication and access control. But, in fact, some of these methods implies many problems. For example, the password is so easy to forget that people can easily be stolen or broken. People with the necessary interaction between the machine safe and convenient transactions, namely the need for simple and quick to use the machine without worrying about the problem of insecurity. Therefore, there is an urgent need for accurate, convenient and safe recognition technology to replace the existing identification.Biometric technology uses the body's physiological or behavioral characteristics for identification, one of the most concern is the fingerprint recognition technology. Fingerprints have the following characteristics of the seven aspects of identity recognition technology to become the first choice:one, universality;two, uniquenes; three, permanent; four, can be collected;five, feasible;six, acceptability;seven, good security features, compared with the account number + password, IC cards and other traditional means of identification, the automated fingerprint identification technology is not lost, not forgotten, the only, and invariance, and better security features, such as highlighting the advantages of ease of use. Using this technology, people can identity and strict matching his fingerprints. At the same time, the fingerprint comparison at very convenient, rapid, and can be compared to achieve better monitoring and surveillance operations, and implementation of one to one and more than one comparison. Fingerprint recognition technology for a modern identity has provided an excellent solution.Automated fingerprint identification technology can be generally divided into the following four modules:one, fingerprint acquisition;two, fingerprint treatment;three, Feature extraction;four, fingerprint matching. Fingerprint acquisition is the fingerprint ridge input distribution by the related equipment, digital; fingerprint pre-processing of fingerprint images collected for a certain treatment, such as fingerprint image segmentation, enhancement, binarization, thinning, etc., making the image after treatment simpler and more easy to extract features; Fingerprint feature extraction after pre-treatment in the image on the basis of the key features of the fingerprint extraction; fingerprint matching is to determine whether two fingerprints from one finger of same people.At present, many companies have participated in the development of fingerprint recognition technology, applications, but most are imported into the fingerprint identification module for system integration, high cost and difficulty in marketing, only a few enterprises have their own algorithm, but Product price is high, the performance of instability, and its cost can not meet the needs of the market Therefore the development of suitable conditions of our country and with international standards and reliable performance, low-cost fingerprint recognition system will cause important innovations for our country, 's anti-counterfeiting technology, identification technology, especially IT industry.VeriFinger is a fingerprint recognition algorithm software which was developed by Net Technology Corporation in 1998. The software version has so far been 12 times to upgrade until the latest version of the current, VeriFinger 6.0, the performance of Maria. At present, VeriFinger SDK has two versions for the current market: one. VeriFinger SDK for the current market, there are two versions: one, VeriFinger 6.0 Standard Edition SDK, it is PC-based fingerprint recognition applications, which includes two main components, namely: "Matcher" and " Extractor";two, VeriFinger 6.0 Extended Edition SDK, primarily for Web and network-based application development environment, which includes the standard version of the entire contents of the SDK also includes client application programming examples and guidance, as well as more than off-the-shelf server (ready - to-use matching server).VeriFinger uses the way of feature points for comparion. VeriFinger development package has the following advantages in fingerprint identification:One:Complete MINEX certification. VeriFinger 6.0 technology is based on MegaMatcher 2.0 fingerprint recognition engine, and it received NIST certification for personal identity authentication applications.Two:Reliablity. Fingerprints at the previous algorithm competition on many occasions, such as FVC2000/2002/2004, as well as the NIST organization FpVTE 2003, even in our earlier version of VeriFinger fingerprint algorithm also achieved good results - we have the performance and accuracy at five name of. VeriFinger 6.0 version of the current in the reliability of the earlier editions have been making considerable progress.Three:Fingerprint matching speed. Get the highest time from many recognition algorithm competition. The time of fingerprint sign in is 0.2-0.4 seconds, at 1:N identification mode to match 40000 fingerprints per-second.Four:VeriFinger can handle rolling fingerprint restrain plane and restrain plane fingerprint, and can compare between the fingerprints at two different acquisition mode .And have very high reliability.This paper describes the specific functions of development package of VeriFinger SDK:including its activation method, feature extraction, feature summary, as well as fingerprint identification of the main function and the use case by C# language. Then, we accomplish the secondary development platform for the automatic fingerprint comparion.On the basis of experiment, we analysis the result of our experiment , and calculate the refuse contractor rates, misjudgment rates and algorithm execution time of fingerprint identification; and then compare the original laboratory results with the results of fingerprint identification system we implemented. Finally, wo get the difference of two different systems, as well as get the advantages of the secondary development system of VeriFinger SD fingerprints.The model also exist a number of shortcomings, VeriFinger SDK various interface functions need further study, the system matches the accuracy and execution time have to be further enhanced, the next step should be to further improve its functionality and performance. |
PDF Full Text Request