Fingerprint Image Segmentation And To Promote Research

With the rapid development of computer technology, automatic fingerprint identification system, which hold a very broad future appliance, has become the hotspot of the biometric recognition field both home and abroad. The modern fingerprint identification technology can accomplish everyone's identity attestation in a very little time, and is used widely in police security, justice, bank, stock trade, examinees' identity attestation, attendance checking , and so on.The fingerprint identification has carried out the century revolution in the field of identity module, and is replacing the traditional security systems which based on token or password.We summarize the common algorithm on the topic of fingerprint image segmentation and enhancement, emphasis on the following and do some improvements.1) Fingerprint image segmentation: Combining the regularity of fingerprint images, we improve the CluD algorithm which makes the segmentation relatively more exact and faster. In addition, this paper realizes a segmentation algorithm based on the plural Gabor filter. Then we further compare the above-mentioned two algorithms.2) Core point detection: We design a new model using a multi-resolution cirque to calculate the core orientation. When detecting the core point, we use such two algorithms as Poincare index and hierarchical analysis of the orientation coherence and then we compare the two algorithms.3) Fingerprint image enhancement: In this paper, we design a new hybrid filter to speed up the fingerprint image enhancement and hold the Fingerprint image texture better. Besides, we improve the spectrum method which is used to estimate the frequency of fingerprint image texture to speed up the calculating and make the estimated frequency more exact. To compare with the result of the hybrid filter, the orientation filter, Fourier filter and Gabor filter for enhancing the fingerprint images are used.