Research On Three-level Fingerprint Classfication System

In recent years, with the rapid development of computer science and network technique, the technology of Automatic fingerprint identification has made considerable progress and has been widely used in finance, police and other fields of people's daily life. However, fingerprint database has become more and more. Automated fingerprint identification system is facing the question about how to retrieval fingerprint quickly in a huge fingerprint database and how to limit scope of the huge fingerprint database in order to reduce the number of fingerprint matching, thereby enhancing the efficiency of fingerprint retrieval. This is the work of fingerprint classification, which has been an important composing of automated fingerprint identification system.We design a three-level Fingerprint Classification System based on the analysis of the basic principles and methods of Fingerprint Classification. By the multi-level classification, it greatly enhanced the efficiency of the Fingerprint Classification. The main work and contributions of this thesis can be summarized as follows:1. First introducing the basic principles of fingerprint classification and some basic knowledge of image pre-processing method.2. Improving the classical formula to compute the poincare index and the algorithm for the singularity detection, so that the modified version of poincare Index can locate the singularities quickly and exactly, and it can improve the speed and accuracy of Fingerprint Classification.3. By the understanding of the general features of fingerprint,in order to classify fingerprint,an innovative way to propose that regarding the degree of angle which formed by connecting between the singularity as a new general characteristics of the fingerprint, and using it in our three-level Fingerprint Classification System.4. Designing a three-level Fingerprint Classification System, which using the number and mutual location of singularities, the degree of angle which formed by connecting between the singularity, Ridge Count as general characteristics of the fingerprint to classify fingerprint.Finally, through by system penetration coefficient, classification accuracy and classification efficiency, performance analysis of three-level Fingerprint Classification System is given out. The results show that the three-level Fingerprint Classification System fully meets the classification requirements of large fingerprint database.