| The fingerprint is unique and stability , and more reliable than other biometrics, so it is used to identify individuals as the main basis for identity. Although fingerprint recognition already have many areas of research results, and many businesses have been developed molding products, it still have far-reaching significance, because these core technologies can not disclose, as well as society has set higher request.On the basis of research results of our country and abroad in recent years in this field, my research mainly includes three parts: fingerprint image Pretreatment, feature extraction and matching fingerprints.Fingerprint image Pretreatment mainly focused on removing the noise fingerprint image processing to enhance image clarity. Pretreatment includes gray standardized work, the calculation of the direction of the ridge line, Gabor filter increased, two-value-treating operation and thinning treatments.Feature extraction extract details of feature points, and then clear the pseudo-removal feature points, and retain real feature point for the work of fingerprint matching well.Fingerprint matching mainly input fingerprint information and the fingerprint template information to compare , and judge whether the two fingerprint images from the same finger. First of all , find the point of reference of the imported image and the template image. And then transform the details of their feature points into polar formula , using their reference point as the polar origin. At last, matching them using variable limits of match boxes.Experiments have been done in the PC of software VC + + 6.0 and experiment results show that the fingerprint recognition algorithm can be more effective and feasible, and can achieve the desired results . |
PDF Full Text Request