An Improved Algorithm for Fingerprint Identification Based on Line Tracking and Mirror-Assisted Method

Abstract

With the continuous development of the society, people's understanding of fingerprint identification technology has also expanded from criminal investigation into commercial field. In recent years, the mainstream fingerprint algorithm is to seek fingerprint feature points on the fingerprint thinning image. These feature points are mainly endpoints and bifurcation points. Although the current fingerprint algorithm achieves high accuracy in both the false acceptance rate (FAR) and the false rejection rate (FRR), it has recently been found that it can be used to unlock other people's fingerprint equipment through the orange peel and the fingerprint sticker. Therefore, the fingerprint algorithm is further studied. This paper makes a great deal of improvement in fingerprint feature extraction and recognition algorithm. The feature points of fingerprint are extracted by using line tracking method and mirror assisted method, which greatly accelerated the extraction time of feature points. Moreover, in the aspect of fingerprint identification, the method of dynamic threshold is adopted to adjust the threshold value according to the number of characteristic points within a certain range in the image to be verified. After each fingerprint is verified successfully, the information extracted from this fingerprint is interacted with the original database, which has achieved the effect of data self-correction. Then, a large number of experiments are carried out to show that the improved fingerprint algorithm can quickly extract feature points and verify identity. And the accuracy of the FAR and the FRR is achieved.