When computer processing fingerprint, only limited fingerprint information is involved, and some calculation errors will be inevitably introduced during the image processing and matching process. therefore, fingerprint algorithm is not the accurate match and the results also can not guarantee 100 % accuracy. In other words, the fingerprint algorithm is not 100% reliable. There are some indicators to assess the performance of a fingerprint algorithm.
FER(Failure To Enroll Rate)
The proportion of the population of applicants failing to complete enrolment.
FRR(False Rejection Rate)
Over a designated number of Genuine Attempts, the proportion of negative answers given by the Biometric System. It is stated as follows:
FRR=NFR/NEIA or FRR=NFR/NEVA
Where NFR (Number of False Rejection): number of negative answers of the Biometric System in the case of Genuine Attempts.
Where NEIA (Number of Enrollee Identification Attempts) is the number of Attempts corresponding to a Genuine Implicit Claim of Identity.
Where NEVA (Number of Enrollee Verification Attempts) is the number of Attempts corresponding to a Genuine Explicit Claim of Identity.
FAR (False Acceptance Rate)
Over a designated number of Impostor Attempts, the proportion of positive answers given by the biometric system. It is stated as follows:
FAR = NFA/NIIA or FAR=NFA/NIVA
Where NFA (Number of False Acceptances) is the number of positive answers of the system in the case of Impostor Attempts.
Where NIIA (Number of Impostor Identification Attempts) is the number of Attempts corresponding to a Impostor Implicit Claim of Identity.
Where NIVA (Number of Impostor Verification Attempts) is the number of Attempts corresponding to a Impostor Explicit Claim of Identity.
EER(Equal Error Rate)
This rate is computed at the point where the FRR Curve and the FAR Curve meet. The EER is then the common value of the two curves at that point. In practice, the two curves are not continuous and a crossover point might not exist. In this case, the interpolated middle of the Equal Error Rate Interval can be reported.
Running times of algorithm in one second.
Code space and data space required by algorithm running.
Because FRR and FAR are contradictory, it is necessary to weigh the ease of use and safety in the application design. An effective approach is to mach two or more fingerprints, and thus greatly enhance the system security while no loss of ease of use at the same time.
The Biometric System Laboratory (University of Bologna), the Pattern Recognition and Image Processing Laboratory (Michigan State University), the Biometric Test Center (San Jose State University) and the Biometrics Research Lab - ATVS (Universidad Autonoma de Madrid) are pleased to announce
FVC2006: the Fourth International Fingerprint Verification Competition
FVC2006 competition focuses on fingerprint verification software assessment. A subset of fingerprint impressions acquired with various sensors was provided to registered participants, to allow them to adjust the parameters of their algorithms. Participants were requested to provide enroll and match executable files of their algorithms; the evaluation was conducted at the organizers?facilities using the submitted executable files on a sequestered database, acquired with the same sensors as the training set.
Miaxis has been participating international Fingerprint Verification Contest(FVC) for three consecutive times since 2002 and ranks top 10 in the world.
Open Category: Medal table
Light Category: Medal table