Fingers Fingerprint Integration And Application Platform Design

Fingerprint Identification Technology is the most widely used identification technology. Biology-based fingerprint recognition system has aroused great interest, and numerous research units and companies are engaged in AFIS (Automated Fingerprint Identification System) research and product development. Nowadays, many products have already entered the business market. But there are still some unresolved problems in the domain of fingerprint identification technology, so it is worth to carry on the study on this field.In this thesis, it firstly analyses the development procedure and status in AFIS, and summarizes the functions of the main parts in AFIS.Based on the discussion, this thesis analyses the normalization methods in the fingerprint pre-processing, then improves the visual result by sub-block normalizing. Then, it discusses two methods of calculating direction, one is mask method and another is formula method, and gives the implementation results of these algorithms. After this, this thesis describes a fingerprint enhancement algorithm, especially for the Gabor filter-based enhancement. Through the results of this algorithm, we could see that these enhancement algorithms are facing the problem between repairing fingerprint image and creating pseudo-details. In the last part of pre-processing part, this thesis discusses and carries out the PCNN-based fingerprint thinning algorithm, and gives the experiment results.In feature extraction and encoding part, this thesis focuses on the FingerCode fingerprint encoding and texture feature extraction method. It discuses and compares this algorithm with the method based on fingerprint minutiae.Then, the thesis gives out a new fingerprint identification technology, which is based on the five-fingerprint data fusion. After this, it analyses the difference between the new technology and the traditional single-fingerprint technology, then lays out the experimental data. After this, it describes an AFIS, in which we use the algorithms discussed in this thesis to form a usable system.