FVC2000: Fingerprint Verification Competition

FVC2000: Fingerprint Verification Competition

2000 | D. Maio, D. Maltoni, R. Cappelli, J. L. Wayman, A. K. Jain
The FVC2000 competition aimed to establish a common benchmark for fingerprint recognition algorithms. Four databases were created using different sensor technologies, and 11 algorithms were tested. The databases included real and synthetic fingerprint images, allowing for a comprehensive evaluation of algorithm performance. The competition was structured to assess the relative performance of algorithms rather than their real-world accuracy. Participants were required to submit two executables for enrollment and matching, with specific performance metrics and time limits. The results showed that some algorithms, such as Sag1 and Sag2, performed well, while others had higher rejection rates. The synthetic database (DB4) was found to be suitable for evaluating algorithm performance, as it produced results similar to the real databases. The competition highlighted the importance of considering both accuracy and efficiency in algorithm design. The results provided a benchmark for future improvements and encouraged ongoing research in fingerprint recognition technology. The FVC2000 initiative aimed to support continued development in this field by providing a standardized testing environment and encouraging the sharing of results.The FVC2000 competition aimed to establish a common benchmark for fingerprint recognition algorithms. Four databases were created using different sensor technologies, and 11 algorithms were tested. The databases included real and synthetic fingerprint images, allowing for a comprehensive evaluation of algorithm performance. The competition was structured to assess the relative performance of algorithms rather than their real-world accuracy. Participants were required to submit two executables for enrollment and matching, with specific performance metrics and time limits. The results showed that some algorithms, such as Sag1 and Sag2, performed well, while others had higher rejection rates. The synthetic database (DB4) was found to be suitable for evaluating algorithm performance, as it produced results similar to the real databases. The competition highlighted the importance of considering both accuracy and efficiency in algorithm design. The results provided a benchmark for future improvements and encouraged ongoing research in fingerprint recognition technology. The FVC2000 initiative aimed to support continued development in this field by providing a standardized testing environment and encouraging the sharing of results.
Reach us at info@study.space