Research And Implementation Of Fingerprint Preprocessing Algorithm

With the rapid development of global information and network, the biological characteristics identification technology, especially the fingerprint identification has been more and more accepted and used in personal identity authentication. The embedded automatic fingerprint identification system has become the key direction of development, because of the low cost, low power consumption, wide development space and many other advantages.Based on the current research achievements of our laboratory, this thesis mainly researches the pre-processing of fingerprint image, and the embedded automatic fingerprint identification system base on small memory. The main work is as follows:1、A new method for fingerprint segmentation base on artificial immune genetic algorithm has proposed. The algorithm combines the immune and genetic mechanism. Through the choice of vaccine and vaccination, the value of the fingerprint image and the antibody concentration are regarded as antibodies affinity function. The algorithm could fast parallel calculate the division of the fingerprint image, and effectively prevent the convergence in local optimal solution. So the algorithm has the high robustness.2、This paper summarized a novel algorithm which base on isotropy method to enhance the fingerprint in recoverable region. The algorithm uses Gabor filtering to enhance fingerprint image in different scales, so it could used in embedded system. This paper also proposes a novel algorithm which combines the advantages of one-dimensional Gabor filtering method and anisotropic method to enhance the fingerprint in recoverable region. The proposed algorithm performs within less time.3、A lower cost embedded automatic fingerprint identification system base on small memory is demonstrated. The embedded system hardware structure, software flow and optimization method is introduced in detail.