|
Fingerprint identification, one of the most extensive applications of biometric technology, with the advantages of the convenience, unique feature, high security, intelligence, and the one-time investment, is the development trend of personal authentication in the future. Now fingerprint identification system widely used in fingerprint time machine, fingerprint access controller, the integration of fingerprint locks, fingerprint safe. Fingerprint recognition technology will be more widely applied to public security organs criminal fingerprint systems, Internet banking, pensioners system in the future. How to miniaturize the fingerprint recognition and increase the accuracy of fingerprint identification is still the focus of the study fingerprint recognition.Identification system for the collection, calculation and corrected the direction of fingerprints, enhanced fingerprints and other aspects have been in-depth focusing in this paper.First of all, point out that direction is the most sensitive and most important factor to the influence of fingerprint image processing. with the analysis software under the VS2005 environment, real-time display each pixel point fingerprints of all parameters, particularly the right direction parameters of real-time displaying and modifying, and verify that direction is the most sensitive and most important factor to the fingerprint image processing.Second, as the direction of the fingerprint core region changes rapidly and largely, point out that correct separately the direction of the core region. On this basis, also point out that we can use the direction of the original picture for entropy way to find the core point, this method can find an effective fingerprint core, this can effectively applies the traditional direction correction to non-core points region, and make it Possibility for separately modifying direction of the core region.Finally, improves the traditional rapidly thinning method, Improved algorithm can effectively prevent the lines shorten phenomenon, provides a guarantee for accurately extract the feature points. The hardware platform verifies that: the three improved methods can effectively prevent noise on the impact of the direction, lay the foundation for the Gabor filter of the enhancement of fingerprint, filter out noise and the restoration of the true fingerprint patterns, and finally extract feature points accurately.
|