Research On Several Fingerprint Image Enhancement

Automated Fingerprint Identification is one of the important Biometrics Recognition, and it is popularly used in the every field of authentication. At the practice, affected by the fingerprint sampler equipment and finger condition, enhancing the gained fingerprint images before extracting minutiae is necessary. Research on fingerprint image enhancement becomes an important part of the Automated Fingerprint Identification System.The present enhancement algorithms mainly have two problems that are too complex and bad robust. This paper mainly research on enhancement methods to solve those problems. For the problems existed in the present methods of estimating fingerprint image's orientation and frequency, a method to estimate orientation and frequency by Radon Transform is proposed, this method can estimate the orientation and frequency correctly to prepare for the Gabor filter enhancement well. At the same time, the Gabor filter enhancement algorithm using the wavelet domain is proposed, this method can restrain the effect of high-frequency disturbance. Since the present methods of orientation filter template have some disadvantages, a whole fingerprint image orientation filter enhancement algorithm is proposed. In this method, filter fingerprint images by new 8 orientations smoothing filter templates and sharpening filter templates, based on the correct block orientations. Besides, for the fuzzy of fingerprint images, a fingerprint image enhancement algorithm combining fuzzy enhancement and orientation filter enhancement methods is proposed. This method reduces the fingerprint images'fuzzy nature by fuzzy enhancement and enhanced the fingerprint images by orientation filter. Experiment results showed that the proposed methods can make good enhancement effect and consume less time.