| Since fingerprint owns uniqueness and life long invariance. Fingerprint Recognition has been one of the most widely used biometrics with the highest recognition rate. Now, it has been applied in a wide range of application fields such as gate control system, attendance management, e-commerce etc. In the general fingerprint identification system, the methods based on minutiae of fingerprints are adopted commonly. To overcome the shortcomings of minutiae-based methods, the structure-based fingerprint algorithm is studied in this thesis. In the structure-based fingerprint algorithm a new fingerprint description system is introduced, which used a Gabor filerbank to extract the texture features so not only local information but also global information are included. Finally, the Euclid Distance between two fmgercode vectors is calculated to implement matching fingerprints.1. The feature extraction method of the fingerprint image segmentation technology is studied. Based on the grayscale difference between two orthogonal directions, a new fingerprint image segmentation algorithm is proposed, and the experimental results show its efficiency.2. Based on the complex filter method, we proposed a new reference point locating algorithm. Designed two direction field calculated algorithm. The experimental results proved that this algorithm has a strong robustness to low quality fingerprints.3. Based on the structure-based fingerprint recognition algorithm, we designed and developed a fingerprint identification software system which included the fingerprint collection module, fingerprint segmentation module, reference point location module, feature extraction module, template registration module and feature matching module.4. The structure-based fingerprint recognition software system is tested on the FVC2000 fingerprint database and the U.are.U 4000 Sensor fingerprint database. The experimental results show that the system has a good rate of recognition, favorable stability and robustness. |
PDF Full Text Request